Data fusion method and device, electronic equipment and storage medium

Abstract

The invention discloses a data fusion method and device, electronic equipment and a storage medium, and relates to the technical field of data processing, due to the fact that query requests of multiple heterogeneous data sources can be received and data retrieval can be carried out based on a dynamic knowledge graph, the limitation of a static knowledge graph is broken through, and real-time fusion and dynamic updating of multi-source data can be achieved; meanwhile, by means of a context segmentation mechanism and a caching strategy in the preset response format, changes of an enterprise organization structure and employee requirements can be flexibly perceived and dynamically managed, and the recommendation strategy is prevented from being rigid, so that the recommendation efficiency is improved. The technical problems of graph information lagging and customer satisfaction caused by an existing architecture of combining a static knowledge graph with a rule engine can be solved, the timeliness and accuracy of job domain data are guaranteed, an insurance product recommendation strategy is more flexible and adaptive, the accurate matching degree between insurance products and enterprise and employee requirements is improved, and the user experience is improved. And the customer satisfaction is improved.

Description

Technical Field

[0001] This disclosure relates to the field of data processing technology, and in particular to a data fusion method and apparatus, electronic device and storage medium. Background Technology

[0002] With the continuous advancement of enterprise digital transformation, occupational insurance management, as an important business model in the life insurance industry, is widely used in group protection services for corporate clients and their employees. Among related technologies, a multimodal data fusion technical system has been constructed through the collaborative operation of model context protocols, graph computing, and dynamic knowledge graphs.

[0003] Existing career domain management methods directly adopt an architecture that combines static knowledge graphs with rule engines, without achieving real-time fusion and dynamic updates of multi-source data. This may lead to outdated graph information and rigid recommendation strategies, making it unable to adapt to rapid changes in corporate organizational structure and employee needs, thereby affecting the accurate matching of insurance products and customer satisfaction. Summary of the Invention

[0004] This disclosure provides a data fusion method, apparatus, electronic device, and storage medium. Its primary purpose is to address the problem of inability to adapt to rapid changes in corporate organizational structures and employee needs, thereby affecting the accurate matching of insurance products and customer satisfaction.

[0005] According to a first aspect of this disclosure, a data fusion method is provided, comprising: Receive job domain data query requests from multiple heterogeneous data sources; Parse the job domain data query request and generate a structured query statement; Based on the structured query statement, matching job domain data is retrieved from the dynamic knowledge graph, and the retrieval results are encapsulated into a preset response format; Through the context segmentation mechanism and caching strategy in the preset response format, real-time collaborative perception of cross-system data and dynamic management of context information are achieved.

[0006] Optionally, receiving job domain data query requests from multiple heterogeneous data sources includes: The request is encapsulated into a preset protocol format that includes a session identifier, query content, context information, and priority. The context information includes the enterprise industry attribute, data source system identifier, and historical query records. The priority field is used to optimize request scheduling during transmission, prioritizing high-priority requests.

[0007] Optionally, parsing the job domain data query request and generating a structured query statement includes: The preset protocol request is transmitted to the target server, which parses the request and calls the LLMs service to generate a structured query statement or a graph traversal instruction. The LLMs service is used to generate the structured query statements and extract job domain tags from unstructured text.

[0008] Optionally, the step of retrieving matching job domain data from the dynamic knowledge graph based on the structured query statement and encapsulating the retrieval results into a preset response format includes: The search results are encapsulated into a preset response format that includes data fields, response status, processing time, and tracking identifiers, and a block return mechanism for large result sets is supported. The dynamic knowledge graph includes real-time graph nodes and a graph neural network model to support multi-hop relation reasoning and semantic association analysis.

[0009] Optional, also includes: In response to changes in the enterprise's organizational structure, a partial update to the dynamic knowledge graph is performed.

[0010] According to a second aspect of this disclosure, a data fusion apparatus is provided, comprising: The receiving unit is used to receive job domain data query requests from multiple heterogeneous data sources; The parsing unit is used to parse the job domain data query request and generate a structured query statement; The retrieval unit is used to retrieve matching job domain data from the dynamic knowledge graph based on the structured query statement, and encapsulate the retrieval results into a preset response format; The management unit is used to realize real-time collaborative perception of cross-system data and dynamic management of context information through the context segmentation mechanism and caching strategy in the preset response format.

[0011] Optionally, the receiving unit is further configured to: The request is encapsulated into a preset protocol format that includes a session identifier, query content, context information, and priority. The context information includes the enterprise industry attribute, data source system identifier, and historical query records. The priority field is used to optimize request scheduling during transmission, prioritizing high-priority requests.

[0012] Optionally, the parsing unit is also used for: The preset protocol request is transmitted to the target server, which parses the request and calls the LLMs service to generate a structured query statement or a graph traversal instruction. The LLMs service is used to generate the structured query statements and extract job domain tags from unstructured text.

[0013] Optionally, the retrieval unit is further configured to: The search results are encapsulated into a preset response format that includes data fields, response status, processing time, and tracking identifiers, and a block return mechanism for large result sets is supported. The dynamic knowledge graph includes real-time graph nodes and a graph neural network model to support multi-hop relation reasoning and semantic association analysis.

[0014] Optional, also includes: The update unit is used to perform a partial update of the dynamic knowledge graph in response to changes in the enterprise's organizational structure.

[0015] According to a third aspect of this disclosure, an electronic device is provided, comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect above.

[0016] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are configured to cause the computer to perform the method described in the first aspect above.

[0017] According to a fifth aspect of this disclosure, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method described in the first aspect above.

[0018] The data fusion method, apparatus, electronic device, and storage medium disclosed herein, through which multiple heterogeneous data source query requests can be received and data retrieval based on dynamic knowledge graphs, overcoming the limitations of static knowledge graphs, and enabling real-time fusion and dynamic updates of multi-source data; simultaneously, by leveraging the context segmentation mechanism and caching strategy in the preset response format, it can flexibly perceive and dynamically manage changes in enterprise organizational structure and employee needs, avoiding rigid recommendation strategies. Therefore, it can solve the technical problems caused by the existing architecture of static knowledge graphs combined with rule engines, such as lagging graph information and rigid recommendation strategies, which cannot adapt to rapid changes in enterprise organizational structure and employee needs, thus affecting the accurate matching of insurance products and customer satisfaction. It achieves the technical effects of ensuring the timeliness and accuracy of occupational domain data, making insurance product recommendation strategies more flexible and adaptable, improving the accuracy of matching insurance products with enterprise and employee needs, and improving customer satisfaction.

[0019] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0020] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein: Figure 1 This is a flowchart illustrating a data fusion method provided in an embodiment of the present disclosure; Figure 2 This is a schematic diagram of the structure of a data fusion device provided in an embodiment of the present disclosure; Figure 3 This is a schematic diagram of another data fusion device provided in an embodiment of the present disclosure; Figure 4 A schematic block diagram of an example electronic device provided for embodiments of this disclosure. Detailed Implementation

[0021] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0022] The data fusion method, apparatus, electronic device, and storage medium of this disclosure are described below with reference to the accompanying drawings.

[0023] Figure 1 This is a schematic flowchart of a data fusion method provided in an embodiment of the present disclosure.

[0024] like Figure 1 As shown, the method includes the following steps: Step 101: Receive job domain data query requests from multiple heterogeneous data sources; Career domain data refers to data related to the expansion of group insurance business for corporate clients. It includes information related to career domain management, such as corporate industry attributes, employee profiles, corporate organizational structure, and employee welfare needs. Heterogeneous data sources cover different types of data sources with different interface protocols and data formats, such as internal HR systems, CRM systems, financial systems, and external publicly available financial reports.

[0025] To effectively receive query requests from these heterogeneous data sources, the system's data layer includes a data adapter. This adapter adapts to the interface specifications and data formats of different heterogeneous data sources, ensuring that various job domain data query requests, such as employee information queries from the HR system, customer policy queries from the CRM system, and enterprise industry data queries from external public data sources, can be received stably and accurately. During the reception process, the system only receives the query request and ensures initial format compatibility; it does not parse or process the specific data content in the request. This lays the foundation for subsequently converting the query request into a format that the system can process and carrying out subsequent data processing, thus meeting the needs of receiving multi-source job domain data query requests in the life insurance job domain operation scenario.

[0026] Step 102: Parse the job domain data query request and generate a structured query statement; Job domain data query requests originate from multiple heterogeneous data sources and may include unstructured or semi-structured query requirements, such as queries about employee welfare policies of companies in a specific industry or queries about the risk levels of employees in a company. The format and expression of these requests will vary depending on the type of data source.

[0027] The system's computation layer is equipped with a dedicated parsing module. This module first identifies key information in job domain data query requests, such as the type of enterprise, employee data dimensions, or insurance business relevance. Then, it organizes and standardizes this key information according to preset parsing rules. During the parsing process, the module adapts to the characteristics of different heterogeneous data source requests. For example, requests from HR systems may contain specific field identifiers, while requests from external public data sources may express the query intent in natural language. The parsing module can accurately extract the core query requirements in all cases.

[0028] After parsing, the system generates structured query statements based on the analyzed query requirements. These statements have a standardized format and conform to data query specifications, making them adaptable to subsequent data query operations. For example, they can be used with SQL statements for relational data queries or graph traversal commands for graph data queries. The entire process revolves solely around parsing job domain data query requests and generating structured statements, without involving data query execution or other subsequent processing steps. This provides standardized query command support for accurately obtaining job domain data and meets the efficient processing needs of multi-source query requests in life insurance job domain operations.

[0029] Step 103: Based on the structured query statement, retrieve matching job domain data from the dynamic knowledge graph, and encapsulate the retrieval results into a preset response format; Based on the structured query statement, matching job domain data is retrieved from the dynamic knowledge graph, and the retrieval results are encapsulated into a preset response format. The dynamic knowledge graph is a structured semantic network storing job domain-related entity attributes and relationships, containing nodes such as enterprise employees, job domain tags, and insurance products. It responds in real-time to changes in enterprise organizational structure and employee needs through an event-driven update mechanism, ensuring the timeliness of stored data. The system's computation layer is configured with a graph computing engine. After receiving the structured query statement, this engine performs node traversal and relationship matching in the dynamic knowledge graph based on the query conditions in the statement, filtering out job domain data that meets the conditions, such as filtering employee information and recommended insurance product data corresponding to specific industry enterprises based on query conditions.

[0030] After the retrieval is completed, the system will encapsulate the acquired job domain data according to a preset response format. The preset response format is a unified data format defined in advance by the system, which covers modules such as data content and status information, to ensure that the encapsulated retrieval results can meet the needs of subsequent transmission and use. The entire process focuses only on the application data retrieval and result encapsulation of structured query statements, providing accurate and standardized data support for subsequent job domain management-related decisions.

[0031] Step 104: Real-time collaborative perception of cross-system data and dynamic management of context information are achieved through the context segmentation mechanism and caching strategy in the preset response format.

[0032] The preset response format is a unified format predefined by the system to standardize the transmission of search results and related information. The context segmentation mechanism is a functional module within this format used to handle long-session scenarios. It splits context information into multiple independent fragments, assigns a unique identifier to each fragment, and sets an expiration time to ensure that context information can be transmitted in segments as needed. The caching strategy is a storage mechanism adopted by the system to improve data interaction efficiency. It temporarily stores results and context fragments related to frequently used job domain data queries through a caching layer, reducing repetitive data processing operations.

[0033] In actual operation, the context segmentation mechanism can adapt to the interaction characteristics of different heterogeneous systems such as HR systems, CRM systems, and external public data sources, ensuring the contextual coherence of job domain data transmitted between different systems during interaction, thereby achieving real-time collaborative awareness of cross-system data. The caching strategy, based on the expiration time of context fragments, periodically cleans up expired cached content while updating newly added high-frequency data, and works in conjunction with the context segmentation mechanism to adjust context information in real time, completing dynamic management of context information. The entire process revolves solely around the application of the context segmentation mechanism and caching strategy, providing stable support for cross-system data interaction and context management in job domain management scenarios.

[0034] In some embodiments, receiving job domain data query requests from multiple heterogeneous data sources includes: The request is encapsulated into a preset protocol format that includes a session identifier, query content, context information, and priority. The context information includes the enterprise industry attribute, data source system identifier, and historical query records. The priority field is used to optimize request scheduling during transmission, prioritizing high-priority requests.

[0035] The default protocol format is a predefined standard format used by the system to regulate request transmission and processing, ensuring that requests from different heterogeneous data sources can be uniformly identified and processed by the system. The default protocol format includes key information such as session identifier, query content context information, and priority. The session identifier is a unique identifier used to distinguish different query requests, which can avoid confusion between different requests during processing and ensure that the processing flow and result feedback of each request can accurately correspond to the data source or user that initiated the request.

[0036] The query content refers to the job domain-related information requested by the requesting party, such as querying employee insurance enrollment status of a specific company or the suitability of group insurance products for a certain industry. Contextual information includes the company's industry attributes, data source system identifier, and historical query records. The company's industry attributes refer to the industry category to which the company belongs, such as manufacturing, internet, or finance. The data source system identifier clarifies which heterogeneous data source the query request originates from, such as the company's internal HR or CRM system, or an external public data platform. Historical query records are records of previous job domain data queries initiated by the requesting party. This contextual information helps the system more accurately understand the query intent and provides a reference for subsequent processing.

[0037] The priority field is used to optimize request scheduling during request transmission. Different job domain data query requests have different levels of urgency. For example, queries involving sudden insurance needs of enterprise employees have a higher priority than regular enterprise job domain data statistics queries. The system can allocate resources to process high-priority requests first when multiple requests are transmitted and waiting to be processed at the same time, based on the priority field, to ensure that urgent query needs can be responded to in a timely manner and improve the efficiency and rationality of overall request processing.

[0038] In some embodiments, parsing the job domain data query request and generating a structured query statement includes: The preset protocol request is transmitted to the target server, which parses the request and calls the LLMs service to generate a structured query statement or a graph traversal instruction. The LLMs service is used to generate the structured query statements and extract job domain tags from unstructured text.

[0039] The target server is the core processing module of the system's computing layer. It possesses the ability to parse requests in preset protocol formats and can identify key elements such as session identifiers, query content context information, and priority within the request, ensuring an accurate understanding of the job domain data query requirements corresponding to the request. After completing request parsing, the target server invokes the LLMs service, or Large Language Model Service, which performs two core functions in this process. Firstly, the LLMs service generates structured query statements based on the parsed query requirements. These structured query statements are adaptable to relational database data query operations, such as generating SQL statements for querying employee welfare data for specific industries. Secondly, the LLMs service can also generate graph traversal instructions, which are suitable for data retrieval in dynamic knowledge graphs, such as generating instructions for traversing the relationships between enterprise nodes and job domain tag nodes.

[0040] Meanwhile, LLMs services also have the ability to extract job domain tags from unstructured text. Job domain tags are feature identifiers related to job domain operations, including enterprise industry tags such as manufacturing tags and internet industry tags, and employee job risk tags such as high-risk job tags and ordinary job tags. LLMs services can accurately extract these job domain tags from unstructured text such as enterprise PDF contracts and employee requirement description documents, providing support for the subsequent generation of structured query statements or graph traversal instructions that meet the requirements, ensuring that the generated query instructions can accurately match the job domain data query needs.

[0041] In some embodiments, retrieving matching job domain data from a dynamic knowledge graph based on the structured query statement and encapsulating the retrieval results into a preset response format includes: The search results are encapsulated into a preset response format that includes data fields, response status, processing time, and tracking identifiers, and a block return mechanism for large result sets is supported. The dynamic knowledge graph includes real-time graph nodes and a graph neural network model to support multi-hop relation reasoning and semantic association analysis.

[0042] The dynamic knowledge graph comprises real-time graph nodes and a graph neural network model. The real-time graph nodes include enterprise nodes, employee nodes, job domain tag nodes, and insurance product nodes. Enterprise nodes record information such as industry attributes and employee numbers; employee nodes store data such as employee departments, positions, and risk levels; job domain tag nodes label features such as high-risk positions in manufacturing; and insurance product nodes include product types, premiums, and coverage information. The graph neural network model possesses multi-hop relationship reasoning and semantic association analysis capabilities. By analyzing the relationships between nodes, such as the WORK_AT relationship between employees and enterprises and the RECOMMEND relationship between job domain tags and insurance products, it enables multi-dimensional data association retrieval across nodes, ensuring accurate location of information matching structured query statements from massive amounts of job domain data.

[0043] After retrieving the search results, the system encapsulates them into a preset response format. This preset response format includes data fields, response status, processing time, and tracking identifiers. Data fields present the retrieved job domain data, such as company name, number of employees, and recommended insurance products. Response status provides feedback on the search operation result, such as success or failure. Processing time records the time from initiating the search to obtaining the result, facilitating system efficiency evaluation. Tracking identifiers track the entire search process, aiding in subsequent troubleshooting and process optimization. Furthermore, this preset response format supports a chunked return mechanism for large result sets. When the search result data volume is large, the system will split the results into multiple data chunks and return them gradually, avoiding transmission delays or interruptions due to excessive data volume, ensuring efficient transmission and stable acquisition of job domain data search results.

[0044] In some embodiments, it also includes: In response to changes in the enterprise's organizational structure, a partial update to the dynamic knowledge graph is performed.

[0045] Dynamic knowledge graphs are structured semantic networks that store the attributes and relationships of entities related to the job domain. They include enterprise nodes, employee nodes, department nodes, and job domain tag nodes. The nodes are connected by edge relationships. For example, employee nodes and department nodes are connected by affiliation relationships, and department nodes and enterprise nodes are connected by subordination relationships. These nodes and edge relationships together constitute the association network of job domain data, providing a foundation for job domain data query and retrieval.

[0046] In the life insurance job domain management scenario, organizational structure changes are common, such as merging or splitting departments, creating new departments, or adjusting departmental functions. These changes lead to alterations in job domain data, including departmental divisions and employee affiliations. If the relevant data in the dynamic knowledge graph is not updated in a timely manner, it will affect the accuracy of subsequent job domain data queries and retrievals, thereby impacting the rationality of job domain management decisions. Therefore, when the system detects an organizational structure change event, it immediately triggers a partial update process for the dynamic knowledge graph.

[0047] During partial updates, the system first identifies the nodes and edge relationships involved in the changes. For example, when merging departments, it identifies the old department nodes to be merged, the newly established department nodes, and the associated employee nodes. The system then adjusts these nodes and edge relationships, such as transferring employee affiliations from old department nodes to new department nodes, deleting invalid edge relationships from old department nodes, and updating the attribute information of the new department nodes, such as department functions and number of employees. This partial update method avoids a full update of the entire dynamic knowledge graph, adjusting only the data affected by the changes. This ensures the timeliness and accuracy of the dynamic knowledge graph data while reducing system resource consumption and improving update efficiency. It ensures that subsequent job domain data retrieval and job domain management decisions based on the dynamic knowledge graph can be conducted using accurate enterprise organizational structure information.

[0048] Corresponding to the data fusion method described above, this invention also proposes a data fusion apparatus. Since the apparatus embodiments of this invention correspond to the method embodiments described above, details not disclosed in the apparatus embodiments can be referred to in the method embodiments described above, and will not be repeated here.

[0049] Figure 2 This is a schematic diagram of the structure of a data fusion device provided in an embodiment of the present disclosure, as shown below. Figure 2 As shown, it includes: The receiving unit 21 is used to receive job domain data query requests from multiple heterogeneous data sources; Parsing unit 22 is used to parse the job domain data query request and generate a structured query statement; The retrieval unit 23 is used to retrieve matching job domain data from the dynamic knowledge graph based on the structured query statement, and encapsulate the retrieval results into a preset response format; The management unit 24 is used to realize real-time collaborative perception of cross-system data and dynamic management of context information through the context segmentation mechanism and caching strategy in the preset response format.

[0050] Furthermore, in one possible implementation of this disclosure, the receiving unit 21 is further configured to: The request is encapsulated into a preset protocol format that includes a session identifier, query content, context information, and priority. The context information includes the enterprise industry attribute, data source system identifier, and historical query records. The priority field is used to optimize request scheduling during transmission, prioritizing high-priority requests.

[0051] Furthermore, in one possible implementation of this disclosure, the parsing unit 22 is further configured to: The preset protocol request is transmitted to the target server, which parses the request and calls the LLMs service to generate a structured query statement or a graph traversal instruction. The LLMs service is used to generate the structured query statements and extract job domain tags from unstructured text.

[0052] Furthermore, in one possible implementation of this disclosure, the retrieval unit 23 is further configured to: The search results are encapsulated into a preset response format that includes data fields, response status, processing time, and tracking identifiers, and a block return mechanism for large result sets is supported. The dynamic knowledge graph includes real-time graph nodes and a graph neural network model to support multi-hop relation reasoning and semantic association analysis.

[0053] Furthermore, in one possible implementation of the embodiments of this disclosure, such as Figure 3 As shown, it also includes: Update unit 25 is used to perform a partial update of the dynamic knowledge graph in response to changes in the enterprise's organizational structure.

[0054] It should be noted that the foregoing explanation of the method embodiments also applies to the apparatus of the embodiments of this disclosure, and the principle is the same. Therefore, the embodiments of this disclosure are not limited thereto.

[0055] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0056] Figure 4 A schematic block diagram of an example electronic device 400 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0057] like Figure 4As shown, device 400 includes a computing unit 401, which can perform various appropriate actions and processes based on a computer program stored in ROM (Read-Only Memory) 402 or a computer program loaded from storage unit 408 into RAM (Random Access Memory) 403. RAM 403 may also store various programs and data required for the operation of device 400. The computing unit 401, ROM 402, and RAM 403 are interconnected via bus 404. I / O (Input / Output) interface 405 is also connected to bus 404.

[0058] Multiple components in device 400 are connected to I / O interface 405, including: input unit 406, such as keyboard, mouse, etc.; output unit 407, such as various types of monitors, speakers, etc.; storage unit 408, such as disk, optical disk, etc.; and communication unit 409, such as network card, modem, wireless transceiver, etc. Communication unit 409 allows device 400 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0059] The computing unit 401 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 401 include, but are not limited to, CPUs (Central Processing Units), GPUs (Graphics Processing Units), various special-purpose AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSPs (Digital Signal Processors), and any suitable processor, controller, microcontroller, etc. The computing unit 401 performs the various methods and processes described above, such as data fusion methods. For example, in some embodiments, the data fusion method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 408. In some embodiments, part or all of the computer program may be loaded and / or installed on device 400 via ROM 402 and / or communication unit 409. When the computer program is loaded into RAM 403 and executed by the computing unit 401, one or more steps of the methods described above may be performed. Alternatively, in other embodiments, the computing unit 401 may be configured to perform the aforementioned data fusion method by any other suitable means (e.g., by means of firmware).

[0060] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application-Specific Standard Products), SOCs (System-on-Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0061] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0062] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, EPROM (Electrically Programmable Read-Only Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0063] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0064] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include LANs (Local Area Networks), WANs (Wide Area Networks), the Internet, and blockchain networks.

[0065] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service system that addresses the shortcomings of traditional physical hosts and VPS (Virtual Private Server) services, such as high management difficulty and weak business scalability. Servers can also be servers for distributed systems or servers incorporating blockchain technology.

[0066] It's important to note that artificial intelligence (AI) is the study of enabling computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It encompasses both hardware and software technologies. AI hardware technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, and big data processing. AI software technologies primarily include computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graph technologies.

[0067] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0068] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A data fusion method, characterized in that, include: Receive job domain data query requests from multiple heterogeneous data sources; Parse the job domain data query request and generate a structured query statement; Based on the structured query statement, matching job domain data is retrieved from the dynamic knowledge graph, and the retrieval results are encapsulated into a preset response format; Through the context segmentation mechanism and caching strategy in the preset response format, real-time collaborative perception of cross-system data and dynamic management of context information are achieved.

2. The method according to claim 1, characterized in that, The process of receiving job domain data query requests from multiple heterogeneous data sources includes: The request is encapsulated into a preset protocol format that includes a session identifier, query content, context information, and priority. The context information includes the enterprise industry attribute, data source system identifier, and historical query records. The priority field is used to optimize request scheduling during transmission, prioritizing high-priority requests.

3. The method according to claim 1, characterized in that, The process of parsing the job domain data query request and generating a structured query statement includes: The preset protocol request is transmitted to the target server, which parses the request and calls the LLMs service to generate a structured query statement or a graph traversal instruction. The LLMs service is used to generate the structured query statements and extract job domain tags from unstructured text.

4. The method according to claim 1, characterized in that, The step of retrieving matching job domain data from the dynamic knowledge graph based on the structured query statement and encapsulating the retrieval results into a preset response format includes: The search results are encapsulated into a preset response format that includes data fields, response status, processing time, and tracking identifiers, and a block return mechanism for large result sets is supported. The dynamic knowledge graph includes real-time graph nodes and a graph neural network model to support multi-hop relation reasoning and semantic association analysis.

5. The method according to claim 1, characterized in that, Also includes: In response to changes in the enterprise's organizational structure, a partial update to the dynamic knowledge graph is performed.

6. A data fusion device, characterized in that, include: The receiving unit is used to receive job domain data query requests from multiple heterogeneous data sources; The parsing unit is used to parse the job domain data query request and generate a structured query statement; The retrieval unit is used to retrieve matching job domain data from the dynamic knowledge graph based on the structured query statement, and encapsulate the retrieval results into a preset response format; The management unit is used to realize real-time collaborative perception of cross-system data and dynamic management of context information through the context segmentation mechanism and caching strategy in the preset response format.

7. The apparatus according to claim 6, characterized in that, The receiving unit is also used for: The request is encapsulated into a preset protocol format that includes a session identifier, query content, context information, and priority. The context information includes the enterprise industry attribute, data source system identifier, and historical query records. The priority field is used to optimize request scheduling during transmission, prioritizing high-priority requests.

8. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-5.

9. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-5.

10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method according to any one of claims 1-5.

Images