An intelligent agent data storage method and system for airport computer room supervision

Abstract

The present invention belongs to the technical field of data storage, and discloses an intelligent agent data storage method and system for airport computer room supervision. The method includes: integrating structured data and unstructured data to form a triple entity relationship group, constructing a unified knowledge graphical data structure, and selecting the Neo4J graph database as the backend storage; storing the unified knowledge graphical data structure in the Neo4J graph database, which improves the query speed of extremely large relational supervision data in the airport computer room and facilitates subsequent supervision data analysis and function expansion. In addition, the method provides a unified query interface for the graphical data of the computer room, enabling users to conveniently perform data retrieval and query, and also providing strong support for the refined management and intelligent development of the airport computer room.

Description

Technical Field

[0001] The present invention belongs to the technical field of data storage, and particularly relates to an intelligent agent data storage method and system for airport computer room supervision. Background Art

[0002] With the continuous deepening of the application of informatization and intelligent technologies in airport computer room management, various types of information in airport computer rooms have shown explosive growth. However, traditional data storage methods have limitations in processing structured and unstructured data, resulting in difficulties in unifying the front-end data query methods, low efficiency, and difficulty in meeting the requirements of efficient supervision and intelligent management of airport computer rooms. Therefore, developing an intelligent agent data storage method and system that can efficiently integrate, store, and query airport computer room information is of great significance for improving the management level of airport computer rooms.

[0003] Patent Invention 1, a traditional Chinese medicine gynecology nursing auxiliary decision-making system based on a clinical knowledge graph (Publication No. CN117476218A, Publication Date: January 30, 2024), discloses a traditional Chinese medicine gynecology nursing auxiliary decision-making system based on a clinical knowledge graph, which relates to the technical field of gynecology nursing, including a construction module for obtaining gynecology nursing information, a matching module connected to the construction module, a selection module connected to the matching module for recommending multiple reference gynecology problem information based on a preset similarity, a comparison module connected to the selection module for determining actual gynecology problem information based on a management terminal, a nursing module connected to the comparison module, and an update module connected to the nursing module. This invention can better assist medical staff in making decisions, reducing the work intensity of medical staff while ensuring accuracy. The focus of the solution of the present invention lies in providing a data storage method and system for an intelligent agent for airport computer room supervision based on a knowledge graph, realizing the integration of structured and unstructured data, improving the rapid query and retrieval of data, and being able to expand and customize query and answer according to the actual needs of airport computer rooms, providing strong support for the refined management and intelligent development of airport computer rooms.

[0004] Patent Invention 2, a visualization supervision system for computer rooms (Publication No. CN116743804A, Publication Date: September 13, 2023), discloses a visualization supervision system for computer rooms, which includes a dynamic environment acquisition subsystem, a security monitoring subsystem, an alarm performance configuration acquisition subsystem, an Internet of Things management platform, an alarm processing platform, and a visualization computer room supervision platform. Through multiple subsystems and platforms, it realizes automatic data acquisition and processing, personnel management, alarm analysis, and closed-loop management of operation and maintenance inspections, bringing new ideas to the operation and maintenance management of computer rooms, improving the digital and intelligent levels of computer room operation and maintenance. This system breaks the barriers of each monitoring subsystem in the computer room and realizes the integrated processing of heterogeneous data of each system, laying a foundation for integrated management; through the visualization computer room supervision platform, it solves the problems of low efficiency and lagging inspection results in traditional inspections, being unable to detect problems in time and causing potential hazards; through the integration and penetration of multiple scenarios, it transforms the traditional extensive computer room management method into a lean management mode, realizing the safety of computer room management, digital monitoring, and unmanned operation and maintenance. The key point of this solution is to provide a visualization supervision system for computer rooms, which realizes the integrated processing of heterogeneous data of each system by breaking the barriers of each monitoring subsystem in the computer room, and constructs a visualization supervision function to improve the security of computer room management and realize the digital and paperless operation and maintenance requirements of its lean management. The focus of the present invention lies in the data storage method of the intelligent agent for airport computer room supervision, which realizes the integration of structured data and unstructured data, can greatly improve the rapid query and retrieval of data, and can be extended and customized for query and answer according to the actual needs of the airport computer room.

[0005] Through the above analysis, the problems and defects existing in the prior art are as follows: In the storage scenarios constructed for the computer room supervision system and the intelligent agent in the prior art, there is no targeted transfer and storage of structured data and unstructured data for the intelligent agent of airport computer room supervision. The query and retrieval speed of data is slow, and it cannot be extended and customized for query applications according to the actual needs of the airport computer room, and it cannot provide a theoretical basis for the decision-making of computer room maintenance personnel. Summary of the Invention

[0006] To overcome the problems existing in the related technologies, the disclosed embodiments of the present invention provide a method and system for storing data of an intelligent agent for airport computer room supervision, specifically relating to a method and system for storing data of an intelligent agent for airport computer room supervision based on a knowledge graph, aiming to achieve efficient integration, rapid query, and intelligent management of airport computer room information. Airport computer room supervision involves structured and unstructured data from multiple data sources. Structured data is generally stored through relational databases such as MySQL, Oracle, and SqlServer, and unstructured data is generally stored and queried through document databases such as MongoDB, HDFS, and Elasticsearch. However, when it comes to scenarios where unified retrieval and query of the two types of data are required, separate processing is often needed, resulting in low query efficiency and unsatisfactory data integration and display. Moreover, in the face of new requirements for computer room supervision, data integration and processing need to be carried out again, and it cannot be rapidly expanded.

[0007] The purpose of the present invention is to connect structured data and unstructured data from multiple data sources. After processing, a knowledge graph for the airport computer room is constructed, and the adjusted graphical data structure is uniformly stored in the Neo4J graph database, providing a high-performance unified query and answer interface for external use, facilitating subsequent function expansion according to application scenarios.

[0008] The technical solution is as follows: A method for storing data of an intelligent agent for airport computer room supervision, including:

[0009] S1. Integrate structured data and unstructured data to form a triple entity relationship group, construct a knowledge graph, and optimize it to form a unified graphical data structure of knowledge;

[0010] S2. Store the unified graphical data structure of knowledge in a graph database for providing data representation and query for subsequent data analysis and function expansion;

[0011] S3. Invoke the data stored in the graph database through a unified query and answer interface, and perform external application function expansion according to specific airport computer room application scenarios.

[0012] In step S1, before integrating structured data and unstructured data to form a triple entity relationship group, data collection is performed. Through sensors and monitoring system data collection devices, structured data such as temperature and humidity monitoring, software monitoring alarms, equipment maintenance, and patrol inspections in the airport computer room are collected in real time; meanwhile, unstructured document data such as emergency response manuals, system installation and deployment records, cable adjustment records, and database common problem handling manuals in the airport computer room are obtained through scanning and entry.

[0013] In step S1, to optimize and form a unified knowledge graphical data structure, the structured data collected is pulled from a third-party database by the system, and the original data attribute fields are mapped and divided into entities and relationships according to semantics. Specifically, it includes:

[0014] Take the attribute field alarm device as a node, and other attribute fields as information nodes. The corresponding relationships are divided into 4 categories: belongs to, alarm time, alarm level, and included alarm content. Determine the relationships between entities pairwise in the knowledge graph through the above 4 types of relationships. The corresponding relationship triples are: <alarm device, belongs to, service group>, <alarm device, of, alarm time>, <alarm device, of, severity level>, <alarm device, includes, alarm content>.

[0015] In step S1, forming triple entity relationship groups from unstructured data includes:

[0016] (2.1) Keyword extraction: Clean the unstructured text data, remove stop words, punctuation marks, and special characters; segment the cleaned text into individual words or phrases; use the TextRank algorithm to extract the keyword expressions in the text, as follows:

[0017]

[0018] In the formula, is the weight of a certain word node at the th iteration, is the damping coefficient, with a value of 0.85; is the node and node correlation between, is the node connected to all nodes, is the node from all nodes that can be reached, is a certain word node at the th iteration, is the number of iterations;

[0019] (2.2) Entity recognition: Deduplicate the extracted keyword expressions and mark their parts of speech; use rule-based named entity recognition NER to identify entities in the document, including person names, system names, and device types, as nodes in the graph relationship.

[0020] (2.3) Relationship extraction: After identifying all entities in the document, set a rule template through the system visualization interface;

[0021] (2.4) Combine the identified entities in pairs, determine the specific relationships between the entities and classify them, and use the triples with entity relationships as the output; among them, the expression of the triples with entity relationships is: , respectively represent two entities, is the relationship between the two entities.

[0022] In step (2.3), the rule templates include:

[0023] The expression form of Email is xxxx@xxx.com;

[0024] Use ^\w+([-+.]\w+)@\w+([-.]\w+).\w+([-.]\w+)*$ to match Email addresses.

[0025] In step S1, construct a knowledge graph, including:

[0026] Data integration: Integrate the data entities and relationships after structured and unstructured processing to form a unified entity relationship triple processing pool;

[0027] Graph construction: Define the nodes and edges in the graph structure according to the identified entities and relationships; the nodes include the name, type, attributes, and affiliated document information of the entities; the edges are marked with the type, direction, and weight information of the relationships.

[0028] Graph structure optimization: Evaluate the quality of the graph relationship of the knowledge graph constructed and presented by the visualization interface of the operation and maintenance system to determine whether further adjustment is required; if adjustment is required, adjust the parameters, delete duplicate or redundant nodes and edges, merge similar entities or relationships, and optimize the layout and links of the graph structure; if no adjustment is required, directly output the graphical data.

[0029] In step S2, store the unified knowledge graphical data structure in the Neo4J graph database, and use the graphical storage characteristics of the Neo4J graph database to store various information of the airport computer room in the form of nodes and relationships to form an airport computer room information database.

[0030] In step S3, external application function extension, including: intelligent question answering, knowledge graph display, and troubleshooting.

[0031] Another object of the present invention is to provide an airport computer room supervision intelligent agent data storage system, which implements the airport computer room supervision intelligent agent data storage method, and the system includes:

[0032] A background service module for connecting structured and unstructured data from multiple data sources, through different processes, to construct a unified new data structure; and to perform function expansion according to application scenarios through a unified query and answer interface;

[0033] A system visualization configuration module for providing an interface for manual input, parameter adjustment, quality assessment data analysis, adjustment, result display and operation for the background service module;

[0034] A data storage module for saving the structured parameter information of the system visualization configuration module, including the connection address of the third-party database to be docked and the update frequency of the knowledge graph, into the relational database Mysql, and at the same time for saving the complete knowledge graph data constructed by the background service module into the Neo4J graph database to realize the graphical storage of data.

[0035] Furthermore, the background service module includes a data collection module, a knowledge graph construction module and a query service module;

[0036] The data collection module: is responsible for collecting various types of information in the airport computer room in real time, including using interfaces to connect to each monitoring system, sensors to obtain structured temperature and humidity monitoring, software monitoring alarms, equipment maintenance, inspection data; through the method of input by the system visualization configuration module, obtaining unstructured document data such as the emergency response manual for the airport computer room, system installation and deployment records, line adjustment records, and common problem handling manuals for the database;

[0037] The knowledge graph construction module: is responsible for cleaning and integrating the collected data to construct the data relationship of the complete computer room knowledge graph;

[0038] The query service module: is responsible for providing a unified query API interface externally, supporting function applications to call this interface to quickly query various types of information in the airport computer room.

[0039] Combined with all the above technical solutions, the beneficial effects of the present invention are as follows:

[0040] The present invention selects the Neo4J graph database as the backend storage, which improves the query speed of a large amount of relational supervision data in the airport computer room and facilitates subsequent supervision data analysis and function expansion; the unified query interface for the graphical data of the computer room in the present invention enables users to conveniently perform data retrieval and query, and at the same time provides strong support for the refined management and intelligent development of the airport computer room. The present invention integrates multi-source structured and unstructured data in the airport computer room, transforms it into a unified knowledge graph, selects the Neo4J graph database as the backend storage, improves the query speed of a large amount of relational supervision data in the airport computer room, and facilitates subsequent supervision data analysis and function expansion; the unified query interface for the graphical data of the computer room enables users to conveniently perform data retrieval and query. Brief Description of the Drawings

[0041] The accompanying drawings herein are incorporated into and constitute a part of this specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure;

[0042] Figure 1 is a flowchart of the method for storing intelligent agent data in an airport computer room provided by an embodiment of the present invention;

[0043] Figure 2 is a schematic diagram of the principle of the method for storing intelligent agent data in an airport computer room provided by an embodiment of the present invention;

[0044] Figure 3 is a flowchart of constructing a knowledge graph provided by an embodiment of the present invention;

[0045] Figure 4 is a schematic diagram of the system for storing intelligent agent data in an airport computer room provided by an embodiment of the present invention. Detailed Embodiments

[0046] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following detailed description of the specific embodiments of the present invention will be given with reference to the accompanying drawings. Many specific details are set forth in the following description to facilitate a thorough understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0047] The innovation of the present invention lies in: the present invention integrates multi-source structured and unstructured data in the airport computer room, transforms it into a unified knowledge graphical data structure, selects the Neo4J graph database as the backend storage, improves the query speed of extremely large relational supervision data in the airport computer room, and facilitates subsequent supervision data analysis and function expansion; the unified query interface for the graphical data in the computer room of the present invention enables users to conveniently perform data retrieval and query, and also provides strong support for the refined management and intelligent development of the airport computer room.

[0048] Embodiment 1, as Figure 1 shown, the method for storing intelligent agent data in an airport computer room provided by an embodiment of the present invention includes:

[0049] S1, integrating structured data and unstructured data to form a triple entity relationship group, constructing a knowledge graph, and optimizing it to form a unified knowledge graphical data structure;

[0050] S2, storing the unified knowledge graphical data structure in a graph database for providing data representation and query for subsequent data analysis and function expansion;

[0051] S3. Call the data stored in the graph database through the unified query and answer interface, and perform external application function expansion according to the specific application scenarios of the airport computer room.

[0052] Embodiment 2. As another possible implementation manner of the embodiments of the present invention, a method for storing intelligent agent data for airport computer room supervision is provided. By collecting various types of information of the airport computer room, after data cleaning and integration, it is saved to the backend Neo4J graph database to realize the construction of a refined information database for the airport computer room, thereby improving the graphical query speed of the airport computer room supervision data.

[0053] By analyzing the current situation of the operation and maintenance of the airport computer room, the characteristics of the current scenario are summarized: 1. It is necessary to interface with structured and unstructured data from multiple data sources, and after different processing, construct a unified new data structure; 2. Unify the storage method of the processed new data structure; 3. Provide a unified query and answer interface with relatively high performance to facilitate function expansion according to application scenarios in the later stage. Based on the above characteristics, in this solution, the structured data is subjected to entity field mapping processing, and the unstructured data is subjected to keyword extraction, entity recognition, and relationship extraction, and then integrated with the data after the mapping of the structured data to construct a knowledge graph, forming a unified graphical data structure, which is stored in the Neo4J graph database, and provides a unified query interface and interface to the outside world, improving the query and retrieval speed of the data and making it more convenient for the airport to expand and customize function applications such as query and answer according to actual needs.

[0054] As Figure 2 shown in the figure, the structured data is stored in a third-party database, and after data entity mapping, a unified new data structure triple entity relationship group is formed; after the unstructured data is processed such as keyword extraction, entity recognition, and relationship extraction, a triple entity relationship group is also formed; the system integrates the triple entity relationship groups of the structured data and the unstructured data, constructs a knowledge graph, stores the optimized unified knowledge graphical data structure in the Neo4J graph database, and the system provides a unified query and answer interface to the outside world; according to the specific application scenarios of the airport computer room, such as intelligent question answering, knowledge graph display, troubleshooting, etc., function applications are expanded.

[0055] Embodiment 3. As another possible implementation manner of the embodiments of the present invention, a method for storing intelligent agent data for airport computer room supervision is provided, specifically including:

[0056] (1) Data collection: Through data collection devices such as sensors and monitoring systems, real-time collect structured data such as temperature and humidity monitoring, software monitoring alarms, equipment maintenance, and patrol inspections of the airport computer room;

[0057] Meanwhile, unstructured document data such as the emergency response manual for the airport computer room, system installation and deployment records, line adjustment records, and database common problem handling manuals are obtained through scanning, inputting, etc.

[0058] (2)Data processing: The collected structured data is pulled out from the third-party database through the system, and then the original data attribute fields are mapped and divided into entities and relationships according to semantics. The structured alarm data is shown in Table 1.

[0059] Table 1 Structured Alarm Data

[0060]

[0061] Take the attribute field "alarm device" as a node, and other attribute fields as information nodes. The corresponding relationships are divided into 4 categories: belonging to, alarm time, alarm level, and included alarm content; judge the relationships between entities in the knowledge graph through the above 4 types of relationships, and list the corresponding relationship triples: <alarm device, belonging to, service group>, <alarm device, of, alarm time>, <alarm device, of, severity level>, <alarm device, including, alarm content>. Unstructured data, such as emergency response manuals, line adjustment records, installation manuals, etc., are transformed into a new data structure of graph relationships through a series of processing steps by the system. The specific steps are as follows:

[0062] (2.1)Keyword extraction: Clean the unstructured text data, including removing stop words, punctuation marks, special characters, etc.; then split the cleaned text into individual words or phrases; use the TextRank algorithm to extract keywords in the text, and the expression is:

[0063]

[0064] In the formula, is the weight of a certain word node at the th iteration, is the damping coefficient, with a value of 0.85; is the relevance between node and node ; is all the nodes connected to node ; is all the nodes that can be reached starting from node ; is the weight of a certain word node at the th iteration, is the number of iterations;

[0065] (2.2) Entity Recognition: Deduplicate the extracted keywords and annotate their parts of speech; use rule-based NER (Named Entity Recognition) to identify entities in the document, such as names of people, system names, equipment types, etc., as nodes in the graph relationship.

[0066] (2.3) Relationship Extraction: After identifying all entities in the document, the operation and maintenance experts set rule templates through the visual interface of the operation and maintenance system, for example:

[0067] The representation form of Email is usually xxxx@xxx.com; use “^\w+([-+.]\w+)@\w+([-.]\w+).\w+([-.]\w+)*$” to match Email addresses.

[0068] (2.4) Combine the identified entities in pairs, determine the specific relationships between the entities and classify them. Finally, take the triples with entity relationships as the output, for example: , respectively represent two entities, is the relationship between the two entities.

[0069] (3) Knowledge Graph Construction, as Figure 3 shown, including:

[0070] Data Integration: Integrate the data entities and relationships after structured and unstructured processing to form a unified entity-relationship triple processing pool.

[0071] Construct Knowledge Graph: Define the nodes and edges in the graph structure according to the identified entities and relationships. The nodes contain information such as the name, type, attributes, and the document to which the entity belongs; the edges annotate information such as the type, direction, and weight of the relationship.

[0072] Graph Structure Optimization: The operation and maintenance personnel can conduct quality assessment on the graph relationship of the constructed knowledge graph presented through the visual interface of the operation and maintenance system to determine whether further adjustment is needed. If adjustment is needed, the operation and maintenance personnel can adjust the parameters, delete duplicate or redundant nodes and edges, merge similar entities or relationships, and optimize the layout and links of the graph structure; if no adjustment is needed, the graphical data can be directly output.

[0073] (4) Data Storage: Store the finally optimized unified knowledge graphical data structure in the Neo4J graph database. Utilize the graphical storage characteristics of the Neo4J graph database to store various types of information of the airport computer room in the form of nodes and relationships to form an airport computer room information library.

[0074] (5) Data query and application: The system constructs a unified query API interface for the airport computer room. Maintenance personnel in the computer room can call this API interface through different visual function interfaces to quickly query various types of information in the airport computer room. The system can provide the best solution suggestions for the computer room maintenance personnel based on the query results to assist them in making decisions.

[0075] Example 4, as Figure 4 shown, an embodiment of the present invention provides an intelligent agent data storage system for airport computer room supervision, including a system visualization configuration module, a background service module, and a data storage module.

[0076] The background service module is used to connect structured and unstructured data from multiple data sources, process them differently to construct a unified new data structure; and perform function expansion according to the application scenario through a unified query and answer interface;

[0077] The system visualization configuration module is used to provide a manual input interface for the data acquisition module, adjust the parameters of the knowledge graph construction module, analyze, adjust, and display the results of model quality assessment data analysis, and visually display and operate other configuration contents required by the system;

[0078] The data storage module is used to save the structured parameter information of the system visualization configuration module, including the connection address of the third-party database docked and the update frequency of the knowledge graph, to the relational database Mysql. At the same time, it is used to save the complete knowledge graph data constructed by the background service module to the Neo4J graph database to realize the graphical storage of data.

[0079] Exemplarily, the background service module further includes a data acquisition module, a knowledge graph construction module, and a query service module.

[0080] The data acquisition module: is responsible for collecting various types of information in the airport computer room in real time, including using interfaces to connect to various monitoring systems, sensors to obtain structured temperature and humidity monitoring, software monitoring alarms, equipment maintenance, inspection and other data; obtaining unstructured document data such as airport computer room emergency response manuals, system installation and deployment records, wire adjustment records, and database common problem handling manuals through the input method of the system visualization configuration module.

[0081] The knowledge graph construction module: is responsible for cleaning and integrating the collected data to construct the data relationship of the complete computer room knowledge graph.

[0082] The query service module: is responsible for providing an external unified query API interface to support function applications to call this interface to quickly query various types of information in the airport computer room.

[0083] As can be seen from the above embodiments, the present invention collects various structured and unstructured data in the airport computer room, forms a unified graphical data structure through data mapping, text automatic classification, and data integration, stores it in the Neo4J graph database that is more convenient for retrieval and use, provides a unified query interface and a display in the form of a knowledge graph externally, greatly improves the query and retrieval speed of data, and can expand and customize query applications according to the actual needs of the airport computer room to assist computer room maintenance personnel in decision-making processing.

[0084] Specifically, the present invention proposes a method for processing and storing structured and unstructured data from multiple data sources: including integrating the data after unifying it into a triple entity relationship group, constructing a knowledge graph, and storing it in the Neo4J graph database to ensure the query efficiency of the computer room graphical data. The intelligent agent graphical data unified query and answer interface in the airport computer room supervision is convenient for later function expansion according to the application scenario. Through this method, the integration of multi-source structured and unstructured data in the airport computer room is realized, and the convenience of converting it into a unified knowledge graph is achieved. Selecting the Neo4J graph database as the backend storage makes full use of the advantages of the graph database in providing a great deal of relational data representation and query for further utilization and query of data, not only improving the query speed but also facilitating subsequent data analysis and function expansion. This solution provides a unified query interface for the computer room graphical data, enabling users to conveniently retrieve and query data, and at the same time providing strong support for the refined management and intelligent development of the airport computer room.

[0085] The invention has obtained verification of fast data query in the intelligent question and answer module of a certain airport intelligent operation and maintenance management platform, providing strong support for the refined management and intelligent development of the airport computer room.

[0086] The above is only a relatively optimal specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention shall be covered by the protection scope of the present invention.

Claims

1. A data storage method for an airport computer room supervision agent, characterized in that: This method integrates the multi-source structured and unstructured data of the airport computer room and transforms them into a unified knowledge graphical data structure. The Neo4J graph database is selected as the backend storage to improve the query speed of the airport computer room's extremely large relational regulatory data, which is convenient for subsequent regulatory data analysis and function expansion. Through the unified query interface of the computer room's graphical data, users can perform data retrieval and query, which also provides strong support for the refined management and intelligent development of the airport computer room. The method includes the following steps: S1, integrate structured data and unstructured data into ternary entity relationship groups, build a knowledge graph, and optimize to form a unified knowledge graphical data structure; S2, stores the unified knowledge graphical data structure in a graph database to provide data representation and query for subsequent data analysis and function expansion; S3, calls the data stored in the graph database through the unified query and answer interface, and expands external application functions according to specific airport computer room application scenarios; In step S1, before integrating the structured data and unstructured data into a ternary entity relationship group, data collection is performed. The temperature and humidity monitoring, software monitoring alarm, equipment maintenance, and inspection structured data of the airport computer room are collected in real time through sensors and monitoring system data collection equipment; at the same time, the airport computer room emergency response manual, system installation and deployment records, line adjustment records, and database common problem handling manual unstructured document data are obtained by scanning and inputting; In step S1, the unified knowledge graphical data structure is optimized by pulling the collected structured data from the third-party database through the system, and mapping and dividing the original data attribute fields into entities and relationships according to semantics; specifically, it includes: The attribute field alarm device is regarded as a node, and other attribute fields are regarded as information nodes. The corresponding relationships are divided into four categories: belonging to, alarm time, alarm level, and included alarm content. The above four types of relationships are used to determine the relationship between entities in the knowledge graph. The corresponding relationship triples are: <alarm device, belonging to, business group>, <alarm device, of, alarm time>, <alarm device, of, severity level>, <alarm device, including, alarm content>. In step S1, the unstructured data is formed into a triple entity relationship group, including: (2.1) Keyword extraction: Clean the unstructured text data to remove stop words, punctuation marks, and special characters; split the cleaned text into individual words or phrases; use the TextRank algorithm to extract keywords from the text, the expression is: Where W t+1 (i) is the weight of a word node i at the t+1th iteration, d is the damping coefficient, which is 0.85; w ji is the correlation between node i and node j, In(i) is all nodes connected to node i, Out(j) is all nodes that can be reached from node j, W t (j) is the weight of a word node j at the tth iteration, where t is the number of iterations; (2.2) Entity recognition: Remove duplicates from the extracted keywords and annotate their parts of speech; use rule-based named entity recognition (NER) to identify entities in the document, including names of people, system names, and device types, as nodes in the graph relationship; (2.3) Relationship extraction: After identifying all entities in the document, set the rule template through the system visual interface; (2.4) The identified entities are combined in pairs, the specific relationships between the entities are determined and classified, and the triples with entity relationships are output; the expression of the triples of entity relationships is: <w i , k ij , w j >, w i , w j Represent two entities, k ij is the relationship between two entities; In step (2.3), the rule template includes: Email is in the form of xxxx@xxx.com; Use ^\w+([-+.]\w+)@\w+([-.]\w+).\w+([-.]\w+)*$ to match the email address; In step S1, a knowledge graph is constructed, including: Data integration: Integrate the data entities and relationships after structured and unstructured processing to form a unified entity-relationship triple processing pool; Graph construction: Based on the identified entities and relationships, define the nodes and edges in the graph structure; the nodes contain the entity's name, type, attributes, and document information; the edges annotate the relationship's type, direction, and weight information; Graph structure optimization: perform quality assessment based on the knowledge graph relationships constructed and displayed on the visual interface of the operation and maintenance system to determine whether further adjustments are needed; if adjustments are needed, adjust parameters, delete duplicate or redundant nodes and edges, merge similar entities or relationships, and optimize the layout and links of the graph structure; if adjustments are not needed, directly output the graphical data; In step S2, the unified knowledge graphical data structure is stored in the Neo4J graph database. By utilizing the graphical storage characteristics of the Neo4J graph database, various types of information of the airport computer room are stored in the form of nodes and relationships to form an airport computer room information database.

2. The data storage method for airport computer room supervision agent according to claim 1 is characterized in that: In step S3, the external application functions are expanded to include: intelligent question and answer, knowledge graph display, and troubleshooting.

3. An airport computer room supervision agent data storage system, characterized in that: The system implements the data storage method for an airport computer room supervision agent as claimed in any one of claims 1 to 2, and the system comprises: The backend service module is used to connect structured and unstructured data from multiple data sources, build a unified new data structure after different processing, and expand functions according to application scenarios through a unified query and answer interface; System visual configuration module, used to provide manual input interface, parameter adjustment, quality assessment data analysis, adjustment, result display and operation for the background service module; The data storage module is used to save the structured parameter information of the system visualization configuration module, including the link address of the docked third-party database and the update frequency of the knowledge graph, into the relational database Mysql. It is also used to save the complete knowledge graph data constructed by the background service module into the Neo4J graph database to realize graphical storage of data.

4. The airport computer room supervision agent data storage system according to claim 3 is characterized in that: The backend service module includes a data collection module, a knowledge graph construction module and a query service module; The data acquisition module is responsible for real-time collection of various information from the airport computer room, including using interfaces to connect to various monitoring systems, sensors to obtain structured temperature and humidity monitoring, software monitoring alarms, equipment maintenance, and inspection data; Through the input method of the system visual configuration module, the unstructured document data of the airport computer room emergency response manual, system installation and deployment records, line adjustment records, and database common problem handling manual are obtained; The knowledge graph construction module is responsible for cleaning and integrating the collected data to build a complete computer room knowledge graph data relationship; The query service module is responsible for providing a unified query API interface to the outside world, and supports functional applications to call the interface to quickly query various types of information in the airport computer room.

Images