A cybersecurity threat intelligence modeling method using open source intelligence
Through the network security threat intelligence modeling method of open source intelligence, D3.js and attack graph technology, the network information format difference and data attribute ambiguity are solved, and the unified processing and analysis of network security threat intelligence is realized. It is suitable for domestic and foreign Open source intelligence data.
Patent Information
- Application Number
- CN202311208376.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-09-19
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2043-09-19
AI Technical Summary
In the prior art, network information formats vary greatly, making it difficult to extract uniformly, and the differences in data characteristics attributes lead to attribute alignment and attribute ambiguity, affecting the unified processing and analysis of network security threat intelligence.
Adopting the network security threat intelligence modeling method of open source intelligence, through D3.js' large-scale entity relationship rendering, model parameter calculation and analysis, and attack graph technology, combined with engineering specifications and knowledge graphs, the format of network information is uniformly extracted and extracted. Data unified processing solves the problems of attribute alignment and ambiguity.
It reduces the difference in network information format, realizes unified data processing, improves the analysis efficiency and accuracy of network security threat intelligence, and is suitable for domestic and foreign open source intelligence data.
Smart Images

Figure CN117254953B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of network security technology, and more specifically, to a network security threat intelligence modeling method using open source intelligence. Background Art
[0002] Network security testing involves a large number of testing tools, and each tool has relatively independent functions and configuration management. Taking penetration testing as an example, security testing requires the design of various stages such as port scanning, operating system identification, vulnerability scanning, and vulnerability exploitation. It requires the combined application of various tools to complete effective security testing. In order to improve testing efficiency and reduce testing difficulty, the usual practice is to integrate various tools into the software framework and implement the overall call of various testing tools through a unified interface.
[0003] Solve the problem of describing the attributes of various security threat activities in the process of network security protection, especially how to use a unified model to describe various attack organizations, attack methods, attack means, and attack resources. In fact, it is to form a highly simulated simulation of possible adversary network attack activities;
[0004] On this basis, we need to conduct cyberspace behavior modeling, examine the attack capabilities of different attack activities on different network assets, as well as the possible impacts and harms, grasp their attack trends, and evaluate their attack effects.
[0005] Based on the above, the inventors found that:
[0006] The emergence of network modeling helps improve the quality of security testing. However, it faces some problems in the actual process. The more prominent problems are:
[0007] 1. Based on network asset situation analysis and device policy configuration, the formats of network information vary greatly, making it difficult to extract in a unified manner;
[0008] 2. When extracting data features, due to the large differences in attributes between different data features, compatibility issues such as attribute alignment and attribute ambiguity are prone to occur when unifying data.
[0009] Therefore, in view of this, the existing structure is studied and improved to provide a network security threat intelligence modeling method using open source intelligence, in order to achieve a more practical and valuable purpose. Summary of the invention
[0010] 1. Technical issues to be solved
[0011] In view of the problems existing in the prior art, the purpose of the present invention is to provide a network security threat intelligence modeling method using open source intelligence, which can reduce the format differences of network information and facilitate the unified extraction of the format. At the same time, it makes the results after the network attack consistent, which can facilitate data unification and solve the problems of difficult attribute alignment and attribute ambiguity.
[0012] 2. Technical solution
[0013] To solve the above problems, the present invention adopts the following technical solutions.
[0014] A network security threat intelligence modeling method using open source intelligence, the specific implementation steps of the method are as follows:
[0015] Step S1: Large-scale entity relationship rendering based on D3.js:
[0016] At this point, the JavaScript library is called, using the widely implemented SVG, JavaScript and CSS standards, improved from the earlier Protovis library, to set animations, dynamically change object states or add tooltips to complete user interaction functions;
[0017] Step S2: Calculate and analyze model parameters to verify the accuracy, rationality and applicability of the model:
[0018] The method of ensuring the quality of the ontology by using engineering specifications is to simplify the problem as necessary according to the characteristics of the actual object and the purpose of modeling, and to put forward some appropriate assumptions in precise language. Based on the assumptions, appropriate mathematical tools are used to characterize the mathematical relationship between the variables, establish the corresponding mathematical structure, calculate all the parameters of the model using the acquired data, and compare the model analysis results with the actual situation to verify the accuracy, rationality and applicability of the model.
[0019] Step S3: Select appropriate components of the target network system, abstractly model them and their interactions at an appropriate granularity, and formally describe the target network and vulnerability knowledge base:
[0020] The attack graph-based defense capability calculation technology builds a model for the entire target network and attackers, uses model analysis tools to conduct a comprehensive analysis of the vulnerabilities of the target network system, discovers the relationship between vulnerabilities, and displays all possible attack paths for attackers to exploit different vulnerabilities in the target network and gradually implement attacks to defeat them one by one. It also selects appropriate components of the target network system and abstractly models them and their interactions at an appropriate granularity to formally describe the target network and vulnerability knowledge base.
[0021] Furthermore, in step S1, when performing entity relationship rendering, the specific implementation steps are as follows:
[0022] Step S1.1: knowledge graph path search method;
[0023] Step S1.2: Path search system based on knowledge graph;
[0024] Step S1.3: Obtain the unweighted PR value group of each node in the knowledge graph and calculate the weighted bidirectional PR value;
[0025] Step S1.4: Optimize software algorithms to improve rendering performance;
[0026] Step S1.5: LOD+quadtree hierarchical block division;
[0027] Step S1.6: Rendering batches, groups models of the same state for rendering, and reduces switching of rendering batches;
[0028] Step S1.7: performing a space segmentation algorithm;
[0029] Step S1.8: further select the data that meets the requirements to reduce the amount of rendering data;
[0030] Step S1.9: perform data scheduling to improve rendering speed;
[0031] Step S1.10: Use web worker to avoid blocking the main thread.
[0032] Furthermore, in step S1.1, the specific steps of the search method are as follows:
[0033] Step S1.11: obtaining at least one node pair selected from the knowledge graph, each of the node pairs including two target nodes;
[0034] Step S1.12: Obtain at least one node pair selected from the knowledge graph, each of the node pairs including two target nodes;
[0035] Step S1.13: generating a target node set based on the target node in the at least one node pair;
[0036] Step S1.14: Generate a graph node set based on the nodes in the multiple triples;
[0037] Step S1.15: Based on the multiple triples, obtain at least one n-hop subgraph related to the target node set by performing a Cartesian product operation based on the target node set and the graph node set;
[0038] Step S1.16: Based on the subgraph path in the at least one hopping subgraph, obtain a path within k hops between the two target nodes of the at least one node pair, where k21,1≤n≤k.
[0039] Furthermore, in step S1.2, the path search system includes an acquisition module, a target node set generation module, a graph node set generation module, a subgraph generation module and a path determination module.
[0040] Furthermore, in step S2, when calculating and analyzing the model parameters, the specific implementation steps are as follows:
[0041] Step S2.1: Use UML without fixed thinking and patterns, and use UML graphics to clearly express the system from analysis to implementation and deployment;
[0042] Step S2.2: construct the ontology and determine the usage requirements according to the ontology situation;
[0043] Step S2.3: Based on the construction method of structured data, further combined with the XML language ~D2R, the structured data is converted into RDF knowledge representation and the data is imported into the knowledge graph;
[0044] Step S2.4: Based on the construction method of semi-structured data, the concept and method of pattern learning are adopted to form one or more patterns by analyzing and learning similar pages. At this time, the data in such pages are automatically extracted to avoid most possible errors and obtain high-quality data;
[0045] Step S2.5: Based on the construction method of unstructured data, knowledge and their connections are extracted from unformatted text or web pages through the set ontology, and the entities and their connections are abstractly modeled using the clause structure.
[0046] Furthermore, in step S2.1, the UML diagram is divided into a structure diagram and a behavior diagram;
[0047] The structural diagram includes class diagram, outline diagram, component diagram, composite structure diagram, object diagram, deployment diagram, and package diagram;
[0048] Behavioral diagrams include activity diagrams, use case diagrams, state machine diagrams, and interaction diagrams;
[0049] Interaction diagrams include sequence diagrams, timing diagrams, communication diagrams, and interaction overview diagrams.
[0050] Furthermore, in step S2.2, the determination method is as follows:
[0051] If the ontology can be in a refined state, then the existing ontology can be expanded or modified to avoid a lot of unnecessary development work;
[0052] If the ontology is in a state that cannot meet the current application requirements, the definition information and key information are searched in the ontology to define classes and class inheritance.
[0053] Furthermore, in step S3, the attack graph technology includes an attack graph automatic construction technology and an attack graph-based security analysis technology.
[0054] Furthermore, in step S3, when modeling and describing the vulnerability knowledge base, the specific implementation steps are as follows:
[0055] Step S3.1: Describe the attack pattern knowledge base and formalize it into triples;
[0056] Step S3.2: Use the attack graph modeling language to describe the target environment, and then use it together with the vulnerability knowledge base as input to the attack graph automatic construction engine.
[0057] 3. Beneficial effects
[0058] Compared with the prior art, the advantages of the present invention are:
[0059] ① This solution, based on the network asset situation and equipment strategy configuration, is based on engineering specifications. At this time, feature modeling for actual objects is carried out, and then an architecture diagram is established based on the actual problems during modeling. Data tools are used to establish data relationships between feature variables and further form a data structure. At this time, a unified identifier for the data structure is used to reduce format differences in network information and facilitate unified format extraction;
[0060] ② In this scheme, when extracting features, the network information data is substituted into the attack model based on network security. At this time, the impact of the attack model on network security threats is used to present the data feature attributes as the results under the network attack, avoiding the differences when the attributes are directly compared. At the same time, since the results after the network attack are consistent, it is convenient to unify the data, solving the problem of difficult attribute alignment and attribute ambiguity;
[0061] ③ This solution, when processing data, performs large-scale entity relationship rendering based on D3.js. At this time, the JavaScript library, by establishing a search in the knowledge graph, and by optimizing the algorithm, rendering and space separation, realizes the rendering of data and improves the rendering speed, ensuring the stability of the entire rendering main line, and then ensuring the rendering effect, so that the entire modeling can be displayed more clearly;
[0062] ④ This solution can be applied to domestic and foreign open source intelligence data, and aggregated into a standard format open source intelligence data set to support threat analysis of the target network. BRIEF DESCRIPTION OF THE DRAWINGS
[0063] Figure 1 The modeling process of the present invention is shown in FIG. Figure 1 ;
[0064] Figure 2 The modeling process of the present invention is shown in FIG. Figure 2 ;
[0065] Figure 3 The modeling process of the present invention is shown in FIG. Figure 3 ;
[0066] Figure 4 The modeling process of the present invention is shown in FIG. Figure 4 ;
[0067] Figure 5 This is a key technology relationship diagram of the attack graph technology of the present invention. DETAILED DESCRIPTION
[0068] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention; it is obvious that the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments, and all other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making creative work are within the scope of protection of the present invention.
[0069] Embodiment 1:
[0070] See also Figure 1-5 , a network security threat intelligence modeling method using open source intelligence, the specific implementation steps of the method are as follows:
[0071] Step S1: Large-scale entity relationship rendering based on D3.js:
[0072] At this point, the JavaScript library is called, using the widely implemented SVG, JavaScript and CSS standards, improved from the earlier Protovis library, to set animations, dynamically change object states or add tooltips to complete user interaction functions;
[0073] Step S2: Calculate and analyze model parameters to verify the accuracy, rationality and applicability of the model:
[0074] The method of ensuring the quality of the ontology by using engineering specifications is to simplify the problem as necessary according to the characteristics of the actual object and the purpose of modeling, and to put forward some appropriate assumptions in precise language. Based on the assumptions, appropriate mathematical tools are used to characterize the mathematical relationship between the variables, establish the corresponding mathematical structure, calculate all the parameters of the model using the acquired data, and compare the model analysis results with the actual situation to verify the accuracy, rationality and applicability of the model.
[0075] Step S3: Select appropriate components of the target network system, abstractly model them and their interactions at an appropriate granularity, and formally describe the target network and vulnerability knowledge base:
[0076] The attack graph-based defense capability calculation technology builds a model for the entire target network and attackers, uses model analysis tools to conduct a comprehensive analysis of the vulnerabilities of the target network system, discovers the relationship between vulnerabilities, and displays all possible attack paths for attackers to exploit different vulnerabilities in the target network and gradually implement attacks to defeat them one by one. It also selects appropriate components of the target network system and abstractly models them and their interactions at an appropriate granularity to formally describe the target network and vulnerability knowledge base.
[0077] This security threat intelligence modeling method is based on engineering specifications, extracts object features, models the knowledge base, establishes corresponding mathematical structures, uses the acquired data to calculate all parameters of the model, compares the model analysis results with the actual situation, verifies the accuracy, rationality and applicability of the model, and then builds a model for the entire target network entity and attacker based on the defense capability calculation technology of the attack graph, uses the model analysis tool to conduct a comprehensive analysis of the vulnerabilities of the target network system, reduces the format differences of the information, aligns the attributes of the multi-attribute feature data, and solves the attribute ambiguity problem;
[0078] This security threat intelligence modeling method is suitable for collecting open source intelligence data at home and abroad, and integrating it into a standard format of open source intelligence data set, which is ultimately used to support threat analysis of the target network.
[0079] Embodiment 2:
[0080] Based on the above embodiment 1, further description is given.
[0081] See also Figure 1 and Figure 2 In step S1, when rendering entity relationships, the specific implementation steps are as follows:
[0082] Step S1.1: knowledge graph path search method;
[0083] Step S1.2: Path search system based on knowledge graph;
[0084] Step S1.3: Obtain the unweighted PR value group of each node in the knowledge graph and calculate the weighted bidirectional PR value;
[0085] Step S1.4: Optimize software algorithms to improve rendering performance;
[0086] Step S1.5: LOD+quadtree hierarchical block division;
[0087] Step S1.6: Rendering batches, groups models of the same state for rendering, and reduces switching of rendering batches;
[0088] Step S1.7: performing a space segmentation algorithm;
[0089] Step S1.8: further select the data that meets the requirements to reduce the amount of rendering data;
[0090] Step S1.9: perform data scheduling to improve rendering speed;
[0091] Step S1.10: Use web worker to avoid blocking the main thread.
[0092] In the entire relationship rendering process, one-hop or multi-hop relationships between entities, that is, paths between nodes, are extracted from the knowledge graph. How to efficiently extract paths between nodes in the knowledge graph?
[0093] See also Figure 1 and Figure 2 , in step S1.1, the specific steps of the search method are as follows:
[0094] Step S1.11: Obtain at least one node pair selected from the knowledge graph, each node pair including two target nodes;
[0095] Step S1.12: Obtain at least one node pair selected from the knowledge graph, each node pair including two target nodes;
[0096] Step S1.13: generating a target node set based on the target node in at least one node pair;
[0097] Step S1.14: Generate a graph node set based on the nodes in the multiple triples;
[0098] Step S1.15: Based on the multiple triples, at least one n-hop subgraph related to the target node set is obtained by performing a Cartesian product operation based on the target node set and the graph node set;
[0099] Step S1.16: Based on the subgraph path in at least one hopping subgraph, obtain a path within k hops between two target nodes of at least one node pair, where k21,1≤n≤k.
[0100] See also Figure 1 and Figure 2 In step S1.2, the path search system includes an acquisition module, a target node set generation module, a graph node set generation module, a subgraph generation module and a path determination module.
[0101] An acquisition module is used to acquire at least one node pair selected from a knowledge graph, each node pair including two target nodes; a target node set generation module is used to generate a target node set based on the target nodes in at least one node pair; a graph node set generation module is used to generate a graph node set based on the nodes in a plurality of triples; a subgraph generation module is used to obtain at least one jumping subgraph related to the target node set based on a plurality of triples by performing a Cartesian product operation based on the target node set and the graph node set; a path determination module is used to obtain a path within k hops between two target nodes of at least one node pair based on a subgraph path in at least one jumping subgraph; wherein k21,1≤≤k.
[0102] See also Figure 1 and Figure 3 In step S2, when calculating and analyzing the model parameters, the specific implementation steps are as follows:
[0103] Step S2.1: Use UML without fixed thinking and patterns, and use UML graphics to clearly express the system from analysis to implementation and deployment;
[0104] Step S2.2: construct the ontology and determine the usage requirements according to the ontology situation;
[0105] Step S2.3: Based on the construction method of structured data, further combined with the XML language ~D2R, the structured data is converted into RDF knowledge representation and the data is imported into the knowledge graph;
[0106] Step S2.4: Based on the construction method of semi-structured data, the concept and method of pattern learning are adopted to form one or more patterns by analyzing and learning similar pages. At this time, the data in such pages are automatically extracted to avoid most possible errors and obtain high-quality data;
[0107] Step S2.5: Based on the construction method of unstructured data, knowledge and their connections are extracted from unformatted text or web pages through the set ontology, and the entities and their connections are abstractly modeled using the clause structure.
[0108] See also Figure 1 and Figure 3 ,In step S2.1, the UML diagram is divided into structure diagram and ,behavior diagram;
[0109] The structural diagram includes class diagram, outline diagram, component diagram, composite structure diagram, object diagram, deployment diagram, and package diagram;
[0110] Behavioral diagrams include activity diagrams, use case diagrams, state machine diagrams, and interaction diagrams;
[0111] Interaction diagrams include sequence diagrams, timing diagrams, communication diagrams, and interaction overview diagrams.
[0112] See also Figure 1 and Figure 3 In step S2.2, the determination method is as follows:
[0113] If the ontology can be in a refined state, then the existing ontology can be expanded or modified to avoid a lot of unnecessary development work;
[0114] If the ontology is in a state that cannot meet the current application requirements, the definition information and key information are searched in the ontology to define classes and class inheritance.
[0115] See also Figure 1 and Figure 4 ,In step S3, the attack graph technology includes attack graph automatic construction technology and attack graph-based security analysis technology.
[0116] See also Figure 1 and Figure 4 ,In step S3, when modeling and describing the vulnerability knowledge base, the specific implementation steps are as follows:
[0117] Step S3.1: Describe the attack pattern knowledge base and formalize it into triples;
[0118] Step S3.2: Use the attack graph modeling language to describe the target environment, and then use it together with the vulnerability knowledge base as input to the attack graph automatic construction engine.
[0119] Embodiment 2:
[0120] Based on the above-mentioned Embodiment 1 and Embodiment 2, further description is given.
[0121] When modeling network security threat intelligence using open source intelligence, we first call the JavaScript library, use the widely implemented SVG, JavaScript and CSS standards, and improve the early Protovis library to set animations, dynamically change object states or add tooltips to complete user interaction functions. Then, we use the method of engineering specifications to ensure the quality of the ontology. According to the characteristics of the actual object and the purpose of modeling, we simplify the problem as necessary and put forward some appropriate assumptions in precise language. On the basis of the assumptions, we use appropriate mathematical tools to characterize the mathematical relationship between the variables and establish the corresponding mathematical structure. We use the acquired data to calculate all the parameters of the model and compare the model analysis results with the actual situation to verify the accuracy, rationality and applicability of the model. Finally, based on the defense capability calculation technology of the attack graph, we build a model for the entire target network and the attacker. We use the model analysis tool to conduct a comprehensive analysis of the vulnerabilities of the target network system, find the relationship between the vulnerabilities, and show all possible attack paths for the attacker to exploit different vulnerabilities in the target network and gradually implement attacks to defeat them one by one. We also select appropriate components of the target network system and abstractly model them and their interactions at an appropriate granularity to formally describe the target network and vulnerability knowledge base.
[0122] In this way, the modeling process can be completed, and when modeling, the corresponding mathematical structure is established to verify the accuracy, rationality and applicability of the model, and the description of the vulnerability knowledge base of network security threat intelligence is realized;
[0123] This modeling method can reduce the format differences of network information and facilitate the unified extraction of the format. At the same time, it makes the results after the network attack consistent, which can facilitate data unification and solve the problems of difficult attribute alignment and attribute ambiguity.
[0124] The above is only a preferred specific implementation of the present invention; however, the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical solution and its improved conception within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.
Claims
1. A network security threat intelligence modeling method using open source intelligence, characterized by: The specific implementation steps of this method are as follows: Step S1: Large-scale entity relationship rendering based on D3.js: At this point, JavaScript libraries are called to use the widely implemented SVG, JavaScript, and CSS standards to set animations, dynamically change object states, or add tooltips to complete user interaction functions; Step S2: Calculate and analyze model parameters to verify the accuracy, rationality and applicability of the model: The method of ensuring the quality of the ontology by using engineering specifications is to simplify the problem according to the characteristics of the actual object and the purpose of modeling, and to put forward assumptions in precise language. Based on the assumptions, mathematical tools are used to characterize the mathematical relationship between the variables, and the corresponding mathematical structure is established. All parameters of the model are calculated using the acquired data, and the model analysis results are compared with the actual situation to verify the accuracy, rationality and applicability of the model. Step S3: Select components of the target network system, abstractly model the vulnerabilities and the interaction relationships between the vulnerabilities at a predetermined granularity, and formally describe the target network and the vulnerability knowledge base: Based on the attack graph defense capability calculation technology, a model is built for the entire target network and attackers. The model analysis tool is used to conduct a comprehensive analysis of the vulnerabilities of the target network system, discover the relationship between vulnerabilities, and show all possible attack paths for attackers to exploit different vulnerabilities in the target network and gradually implement attacks to defeat them one by one. The components of the target network system are selected and the vulnerabilities and the interaction relationships between the vulnerabilities are abstractly modeled at a predetermined granularity, formally describing the target network and vulnerability knowledge base; In step S1, when performing entity relationship rendering, the specific implementation steps are as follows: Step S1.1: Search the knowledge graph path; Step S1.2: Search in the system based on the knowledge graph path; Step S1.3: Obtain the unweighted PR value group of each node in the knowledge graph and calculate the weighted bidirectional PR value; Step S1.4: Optimize software algorithms to improve rendering performance; Step S1.5: LOD+quadtree hierarchical block division; Step S1.6: When rendering a batch, models of the same state are grouped for rendering to reduce the switching of rendering batches; Step S1.7: performing a space segmentation algorithm; Step S1.8: further select the data that meets the requirements to reduce the amount of rendering data; Step S1.9: perform data scheduling to improve rendering speed; Step S1.10: Use webworker to avoid main thread blocking; In step S1.1, the specific steps of the search method are as follows: Step S1.11: obtaining at least one node pair selected from the knowledge graph, each of the node pairs including two target nodes; Step S1.12: generating a target node set based on the target node in the at least one node pair; Step S1.13: Generate a graph node set based on the nodes in the multiple triples; Step S1.14: Based on the multiple triples, obtain at least one n-hop subgraph related to the target node set by performing a Cartesian product operation based on the target node set and the graph node set; Step S1.15: Based on the subgraph path in the at least one hopping subgraph, obtain a path within k hops between the at least one node and the two target nodes, where 1≤n≤k.
2. A network security threat intelligence modeling method using open source intelligence according to claim 1, characterized in that: In the step S1.2, the path search system includes an acquisition module, a target node set generation module, a graph node set generation module, a subgraph generation module and a path determination module.
3. The network security threat intelligence modeling method using open source intelligence according to claim 1 is characterized by: In step S2, when calculating and analyzing the model parameters, the specific implementation steps are as follows: Step S2.1: Use UML without fixed thinking and patterns, and use UML graphics to clearly express the system from analysis to implementation and deployment; Step S2.2: construct the ontology and determine the usage requirements according to the ontology situation; Step S2.3: Based on the construction method of structured data, the structured data is further converted into RDF knowledge representation in combination with XML language, and the data is imported into the knowledge graph; Step S2.4: Based on the construction method of semi-structured data, the concept and method of pattern learning are adopted to form one or more patterns by analyzing and learning similar pages. At this time, the data in such pages are automatically extracted to avoid most possible errors and obtain high-quality data; Step S2.5: Based on the construction method of unstructured data, knowledge and its connection with the ontology are extracted from unformatted text or web pages through the set ontology, and the relationship between entities and the ontology is abstractly modeled using the clause structure.
4. A network security threat intelligence modeling method using open source intelligence according to claim 3, characterized in that: In step S2.1, the UML diagram is divided into a structure diagram and a behavior diagram; Structural diagrams include class diagrams, outline diagrams, component diagrams, composite structure diagrams, object diagrams, deployment diagrams, and package diagrams; Behavioral diagrams include activity diagrams, use case diagrams, state machine diagrams, and interaction diagrams; Interaction diagrams include sequence diagrams, timing diagrams, communication diagrams, and interaction overview diagrams.
5. The method for modeling network security threat intelligence using open source intelligence according to claim 3, characterized in that: In step S2.2, the determination method is as follows: If the ontology is in a refined state, then the existing ontology can be expanded or modified to avoid a lot of development work; If the ontology is in a state that cannot meet the current application requirements, the definition information and key information are searched in the ontology to define classes and class inheritance.
6. The network security threat intelligence modeling method using open source intelligence according to claim 1, characterized in that: In step S3, the attack graph technology includes attack graph automatic construction technology and attack graph-based security analysis technology.
7. The method for modeling network security threat intelligence using open source intelligence according to claim 1, characterized in that: In step S3, when modeling and describing the vulnerability knowledge base, the specific implementation steps are as follows: Step S3.1: Describe the attack pattern knowledge base and formalize it into triples; Step S3.2: Use the attack graph modeling language to describe the target environment, and then use it together with the vulnerability knowledge base as input to the attack graph automatic construction engine.
