Cross-industry information system intelligent integration method and system based on knowledge graph

Abstract

The invention discloses a knowledge graph-based cross-industry information system intelligent integration method and system, and relates to the technical field of image processing. The problem of data heterogeneity is solved through the cross-industry data mapping model, the dynamic knowledge graph is constructed in combination with the graph neural network, self-adaptive adjustment is achieved, meanwhile, data interaction safety is guaranteed by means of multiple verification, and the transmission strategy is dynamically adjusted and resource scheduling optimization is achieved. Standardized integration of cross-industry data, dynamic optimization of knowledge association, credible guarantee of interaction security and improvement of transmission efficiency are achieved, and finally intelligent integration and efficient collaboration of a cross-industry information system are achieved.

Description

Technical Field

[0001] This invention relates to the field of underwater vision technology, specifically a cross-industry information system intelligent integration method and system based on knowledge graphs. Background Technology

[0002] In today's era of deep integration of information technology and artificial intelligence, cross-industry knowledge integration and intelligent collaboration have become important drivers of socio-economic development. This field is not only about data interconnection, but also directly impacts the efficient collaboration and innovation capabilities among multiple industries, making its strategic value self-evident. However, current mainstream solutions often reveal their inability to adapt to dynamic environmental changes when facing complex and ever-changing real-world scenarios. In particular, when dealing with the deep correlation of data and knowledge between different industries, they lack sufficient flexibility and accuracy, resulting in low resource utilization efficiency and limited decision support capabilities.

[0003] A deep analysis of the core challenges in this field reveals that the main problem lies in the insufficient dynamic adjustment capability of knowledge structures. Data and knowledge systems across different industries often exhibit high heterogeneity, and existing methods struggle to flexibly adjust the structure and content according to real-time business needs when building and maintaining knowledge connections. This further leads to another critical issue: the credibility and real-time nature of knowledge interaction are difficult to guarantee during multi-domain collaboration, due to the lack of effective mechanisms to ensure the security and consistency of information during its flow. These two factors are closely related; the dynamic adjustment capability of knowledge structures directly affects trust building and efficiency improvement during the interaction process. Failure to address this will severely restrict the realization of cross-industry intelligent collaboration. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for intelligent integration of cross-industry information systems based on knowledge graphs. Through data standardization, dynamic knowledge graph construction and adaptive adjustment, cross-domain trust verification and transmission resource optimization, it can achieve intelligent integration and efficient collaboration of cross-industry information systems.

[0005] The objective of this invention can be achieved through the following technical solutions: This application provides a knowledge graph-based intelligent integration method for cross-industry information systems, comprising the following steps: By establishing a cross-industry data mapping model, the heterogeneous characteristics of the data are standardized. The original datasets are obtained from data sources in multiple industries, and the fields are aligned and semantically labeled according to a unified format to obtain a preliminary integrated data set. Graph neural networks are used to model knowledge structure adjustment. To meet the needs of knowledge association depth and multi-domain data fusion, a dynamic graph of node and edge relationships is constructed to determine the association strength and weight distribution between knowledge points. Obtain real-time feedback data on changes in business scenarios, determine whether the scenario changes indicated by the feedback data exceed a preset threshold, and if they do, trigger an adaptive adjustment mechanism to update the node weights and edge connections in the graph to obtain the adjusted knowledge structure. To address the need for trusted interaction assurance and information security verification, a trust verification protocol for cross-domain data interaction is constructed to obtain identity identifiers and data signatures from both parties involved in the interaction, and to determine the integrity and legality of the interacting data. When the integrity and legality verification of the interactive data pass, the data is allowed to be transmitted in real time within the framework of cross-industry collaborative efficiency. The time delay and packet loss rate data during the transmission process are obtained to determine the degree to which the real-time interaction requirements are met. Obtain relevant indicators for resource utilization optimization, extract the distribution of bandwidth usage and computing load during the transmission process, and determine whether there are resource bottlenecks or uneven distribution. Based on the assessment of resource bottlenecks or uneven distribution, the system adapts to dynamic environments, adjusts data transmission paths and computational task allocation strategies, obtains the latest resource status information, and determines an optimized resource scheduling scheme.

[0006] Furthermore, a preliminary integrated dataset was obtained, specifically including: The original dataset is obtained by cross-industry data sources, and a cross-industry data mapping model is built to perform preliminary classification of the data. The heterogeneous characteristics are standardized. When the data field formats are inconsistent, the format is unified by field alignment tools to determine the dataset with consistent format. For semantic annotation requirements, the data is annotated using preset semantic recognition rules. In combination with industry characteristics and cross-industry requirements, when there is missing or abnormal data, the data completion mechanism is used to determine the complete dataset. The support vector machine algorithm is used to extract features from the data, and data fusion technology is used to integrate the data set construction requirements. Then, semantic disambiguation rules are used to adjust the data with semantic conflicts, resulting in an optimized data set.

[0007] Furthermore, the strength of the connections and weight distribution between knowledge points are determined, specifically including: By integrating the dataset, a graph neural network is used to initially model the knowledge structure, obtain the basic connection relationships between knowledge points, determine the initial node relationships, construct a dynamic graph of edge relationships, calculate the initial association strength between each node, and obtain the structural framework of the dynamic graph. By integrating data from multiple fields and adjusting the deep calculation method of knowledge association, the potential connection strength between each knowledge point is determined. When the potential connection strength is lower than the preset threshold, the edge relationship construction is optimized and adjusted, the association strength is recalculated, and a more accurate weight distribution is obtained. Based on the optimized weight distribution, the node and edge relationships in the dynamic graph are updated to determine the core connection paths in the knowledge structure. Then, the matching degree of knowledge association and multi-domain integration is analyzed, the data set is classified and reorganized, the integration effect of data in each domain is judged, and the final dynamic adjustment result is determined.

[0008] Furthermore, the adjusted knowledge structure is obtained, specifically including: By using a real-time monitoring system, change feedback data is obtained from business scenarios, the correlation strength in the dynamic graph is continuously tracked, and then a comparison analysis is performed using preset thresholds. When the indicator value of the scenario change exceeds the preset threshold, it is determined that an adaptive adjustment mechanism needs to be triggered, and the triggering conditions for adjustment are determined. A preliminary assessment of node weights and edge connections in the dynamic graph is conducted to identify weak links in the current knowledge structure and determine the target regions that need optimization. A graph neural network model is then used to recalculate node weights and edge connections, dynamically adjust the association strength, and obtain updated graph parameters. By updating the graph parameters, the knowledge structure is optimized as a whole, the adjusted node distribution and edge connection status are obtained, the new graph form is determined, and then the results are verified using real-time feedback data from business scenarios. If the verification results show that the scene changes are still outside the preset threshold, the adjustment mechanism is repeated to obtain the final stable knowledge structure and record and store the long-term operating status of the dynamic graph.

[0009] Furthermore, determining the completeness and validity of the interactive data specifically includes: By adjusting the knowledge structure, a verification protocol framework is constructed to address the trust verification requirements in cross-domain interactions. The identity identifiers and data signatures are obtained from both parties involved in the interaction. The identity identifiers are then verified and compared with a pre-established identity database to obtain the identity verification results. Further analysis of the data signature is conducted, and the signature content is decrypted and compared using a digital signature verification tool. The signature verification status is then determined, and the data integrity is checked. Key fields are extracted from the data transmitted between the two parties, and a hash algorithm is used to calculate the data fingerprint to obtain the integrity check conclusion. The system assesses data legality, extracts compliance features from data content, matches them with a pre-defined compliance rule base, determines the legality assessment result, conducts a comprehensive analysis of interaction security, integrates identity verification, signature status, integrity conclusions and legality results, and triggers a security alarm mechanism when any link is abnormal to obtain the final security judgment status. The trust verification protocol for cross-domain interaction is dynamically updated, abnormal data and security alarm information are recorded in the knowledge structure, and the protocol parameters are optimized and adjusted using a graph neural network model to obtain the updated verification framework.

[0010] Furthermore, determine the degree to which real-time interaction requirements are met, specifically including: By analyzing the complete and valid verification results of the interactive data, when the verification result meets the preset standard, the data is triggered to enter the real-time transmission process, obtain the initial transmission status information, and determine whether the transmission is feasible. Using a pre-established transmission channel, real-time transmission operations are performed within a cross-industry business collaboration framework. Time delay data is collected during the transmission process, and packet loss rate information is obtained. By comparing with a preset threshold range, it is determined whether the transmission quality meets the real-time standard. When the transmission quality does not meet the real-time standard, the priority strategy of the transmission channel is adjusted, the time delay and packet loss rate data are collected again, and the support vector machine algorithm is used to comprehensively analyze the time delay and packet loss rate to obtain the evaluation result of the degree of satisfaction of the real-time interaction requirements. By further processing the satisfaction assessment results, if the assessment result is lower than the preset threshold, the backup transmission channel is triggered to reacquire the data indicators during the transmission process and determine the final demand satisfaction status.

[0011] Furthermore, determining whether there are resource bottlenecks or uneven distribution includes: By analyzing data, bandwidth usage distribution and load distribution are extracted from real-time interactive data to obtain resource usage information. When the bandwidth usage distribution in the resource usage information exceeds a preset threshold, the bandwidth usage distribution is obtained again by dynamically adjusting the transmission channel priority to determine the bottleneck relief status. The load distribution is analyzed and calculated. If there is a concentration of load distribution, the load balancing algorithm is used to redistribute the computing tasks to obtain the adjusted load distribution. Then, by comparing the adjusted load distribution with the preset threshold, it is determined whether there is an uneven distribution. If uneven distribution still exists, a clustering algorithm is used to classify and analyze bandwidth usage and computational load, adjust the resource scheduling strategy of the transmission process, obtain new bandwidth usage and computational load distribution, determine the final resolution of resource bottlenecks and uneven distribution, and then use a decision tree algorithm to comprehensively evaluate resource optimization indicators to obtain the resource utilization efficiency for real-time interaction requirements.

[0012] Furthermore, the optimized resource scheduling scheme is determined, specifically including: By periodically collecting the status information of system resources, a resource usage distribution map is generated to identify unevenly distributed or bottleneck nodes. When the status information shows that the resource utilization rate of a certain node exceeds a preset threshold, it is marked as a bottleneck node, and a list of bottleneck nodes is obtained. Graph theory algorithms are used to calculate data transmission paths, generate an optimized set of transmission paths, adjust the allocation ratio of computational tasks, generate a task allocation scheme, and use reinforcement learning algorithms to iteratively update the scheduling strategy according to dynamic environmental changes, determine a dynamic scheduling scheme, and adjust system resource allocation in real time to generate the final resource scheduling result.

[0013] Furthermore, after determining the optimized resource scheduling scheme, it also includes: updating the priority rules for data interaction and task processing, obtaining monitoring data on execution efficiency and response time from the scheduling logs, and judging the improvement of overall collaborative performance.

[0014] This application provides a knowledge graph-based intelligent integration system for cross-industry information systems, used to implement a knowledge graph-based intelligent integration method for cross-industry information systems, including: The data standardization and integration module classifies, unifies the format and semantically labels multi-source heterogeneous raw data through a cross-industry data mapping model, extracts features and fuses data using the support vector machine algorithm, and finally forms a standardized preliminary integrated dataset. The knowledge graph construction and optimization module, based on integrated data, constructs a dynamic knowledge graph through graph neural networks, calculates the association strength and weight distribution of knowledge points, integrates data from multiple domains to mine potential connections, optimizes graph edge relationships, clarifies core connection paths, and performs dynamic adjustments to the knowledge structure and deep integration of data from multiple domains. The adaptive adjustment module monitors the feedback of changes in business scenarios in real time. When the scenario indicators exceed the threshold, the adaptive mechanism is triggered to re-evaluate and calculate the weights and edge connections of the graph nodes, optimize the knowledge structure and verify the effect, and continue to iterate until it is stable. The cross-domain trust verification module constructs a cross-domain interaction trust verification protocol, verifies the identities and data signatures of both parties involved in the interaction, detects data integrity and legality through hash algorithms and compliance rule bases, integrates multi-dimensional verification results to trigger security alarms, and dynamically updates protocol parameters. The real-time transmission and performance evaluation module, after the data verification is passed, realizes cross-industry real-time data transmission through a preset channel, collects latency and packet loss rate indicators to evaluate transmission quality, and adjusts the transmission strategy or activates the backup channel when the standard is not met. The algorithm comprehensively analyzes the data to determine the degree to which the real-time requirements are met. The resource scheduling optimization module analyzes bandwidth usage and computational load data during transmission, generates dynamic scheduling schemes by optimizing resource scheduling in combination with graph theory and reinforcement learning algorithms, and iteratively optimizes collaborative performance based on log data.

[0015] The beneficial effects of this invention are as follows: By using a cross-industry data mapping model, the original data is formatted, semantically labeled, and feature-fused. Combined with support vector machine algorithms and semantic disambiguation rules, the problem of inconsistent formats and semantic ambiguity caused by data heterogeneity is solved. At the same time, a dynamic knowledge graph is constructed using graph neural networks. By calculating the strength of node associations, optimizing edge relationships, and fusing data from multiple domains, the limitations of traditional static knowledge structures are broken. This allows the associations between knowledge points to be dynamically adjusted according to business needs, accurately uncovering deep connections and improving the knowledge system's ability to integrate cross-industry data and its flexibility. By leveraging a real-time monitoring system to track changes in business scenarios, an adaptive mechanism is triggered when indicators exceed thresholds. This mechanism recalculates graph parameters and iteratively optimizes the knowledge structure using a graph neural network, addressing the issue of knowledge structures lagging behind business needs in traditional solutions. This enables real-time matching of the knowledge system with dynamic scenarios, enhancing the system's environmental adaptability. Furthermore, a trust verification protocol is constructed using identity verification, digital signature verification, hash algorithms, and compliance rule matching to comprehensively detect data integrity and legality, eliminating risks such as identity forgery and data tampering during interactions. This establishes a trusted mechanism for cross-industry data interaction, ensuring the secure flow of information. After data verification, transmission latency and packet loss rate data are collected. Support vector machine algorithms are used to evaluate real-time performance and dynamically adjust transmission strategies to solve the problem of insufficient real-time data transmission across industries and ensure efficient data transmission within the collaborative framework. At the same time, by extracting bandwidth usage and computational load data, resource bottlenecks are identified by combining load balancing, clustering, and decision tree algorithms. Graph theory algorithms are used to optimize transmission paths, and reinforcement learning algorithms are used to dynamically schedule resources to solve the problem of uneven resource allocation, significantly improve resource utilization efficiency, and avoid low collaborative efficiency caused by resource imbalance. Ultimately, intelligent integration and efficient collaboration of cross-industry information systems are achieved. Attached Figure Description

[0016] To better understand and implement this application, the technical solution is described in detail below with reference to the accompanying drawings.

[0017] Figure 1 A flowchart illustrating a cross-industry information system intelligent integration method based on knowledge graphs, provided in Embodiment 1 of this application; Figure 2 This is a flowchart illustrating the process of determining the association strength and weight distribution between knowledge points using a knowledge graph-based intelligent integration method for cross-industry information systems, as provided in Embodiment 1 of this application. Figure 3 A flowchart illustrating the adjusted knowledge structure obtained from a knowledge graph-based intelligent integration method for cross-industry information systems provided in Embodiment 1 of this application; Figure 4This is a schematic diagram of the structure of a cross-industry information system intelligent integration system based on knowledge graph, provided in Embodiment 2 of this application. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, exemplary embodiments will be described in detail below, examples of which are illustrated in the accompanying drawings. In the following description relating to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of methods and systems consistent with some aspects of this application as detailed in the appended claims.

[0019] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0020] The following detailed description of the specific implementation methods, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided in detail.

[0021] Example 1

[0022] Please see Figures 1-3 This embodiment provides a cross-industry information system intelligent integration method based on knowledge graphs, including the following steps: S1. By establishing a cross-industry data mapping model, the heterogeneous characteristics of the data are standardized. The original datasets are obtained from data sources in multiple industries, and the fields are aligned and semantically labeled according to a unified format to obtain a preliminary integrated data set.

[0023] Furthermore, a preliminary integrated dataset was obtained, specifically including: The original dataset is obtained by acquiring data from cross-industry data sources, and a cross-industry data mapping model is constructed to perform preliminary classification of the data to obtain a classified dataset. The heterogeneous characteristics are then standardized. When the data field formats are inconsistent, a field alignment tool is used to unify the formats and determine a dataset with consistent formats. For semantic annotation requirements, the data is annotated using preset semantic recognition rules to obtain an annotated dataset. In combination with industry characteristics and cross-industry requirements, if there is missing or abnormal data, a data completion mechanism is used to process it and determine the complete dataset. The support vector machine algorithm is used to extract features from the data, and data fusion technology is used to integrate the data set construction requirements to determine the initial integrated data set. Then, semantic disambiguation rules are used to adjust the data with semantic conflicts to obtain the optimized data set.

[0024] In constructing the cross-industry data mapping model, raw datasets are obtained from multiple industry data sources. Classification rules are designed based on the business attributes and structural characteristics of the data to initially classify the data. Then, a field mapping dictionary is established to address the differences in field formats among different industries, and format standardization is achieved through field alignment tools. At the same time, semantic recognition rules are formulated based on industry characteristics and cross-industry needs to semantically annotate the data and design a data completion mechanism to handle missing or abnormal data. Finally, a complete rule system covering data classification, format mapping, semantic annotation, and anomaly handling is formed to achieve systematic and standardized processing of the heterogeneous characteristics of cross-industry data.

[0025] Specifically, semantic annotation and anomaly handling are completed by combining preset semantic recognition rules and data completion mechanisms. Then, data features are extracted and fused using the support vector machine algorithm. Finally, conflicting data are adjusted through semantic disambiguation rules. This effectively solves the heterogeneity problems such as inconsistent data formats, semantic ambiguity, and missing data across industries. It achieves standardized integration and semantic unification of multi-source data, forming an optimized dataset with standardized structure and accurate semantics. This lays a high-quality data foundation for subsequent knowledge graph construction and cross-industry intelligent integration, and significantly improves the usability of data and the efficiency of cross-domain integration.

[0026] S2. Based on the initially integrated dataset, a graph neural network is used to model the knowledge structure adjustment. To meet the needs of knowledge association depth and multi-domain data fusion, a dynamic graph of node and edge relationships is constructed to determine the association strength and weight distribution between knowledge points.

[0027] Furthermore, the strength of the connections and weight distribution between knowledge points are determined, specifically including: S21. By integrating the dataset, a graph neural network is used to initially model the knowledge structure, obtain the basic connection relationships between knowledge points, determine the initial node relationships, construct a dynamic graph of edge relationships, calculate the initial association strength between each node, and obtain the structural framework of the dynamic graph. S22. Integrate data from multiple fields, adjust the deep calculation method of knowledge association, determine the potential connection strength between each knowledge point, and when the potential connection strength is lower than the preset threshold, optimize and adjust the edge relationship construction, recalculate the association strength, and obtain a more accurate weight distribution. S23. Based on the optimized weight distribution, update the node and edge relationships in the dynamic graph, determine the core connection paths in the knowledge structure, analyze the matching degree of knowledge association and multi-domain integration, classify and reorganize the data set, judge the integration effect of data in each domain, and determine the final dynamic adjustment result.

[0028] Specifically, by using graph neural networks to model the knowledge structure of initially integrated data, it is possible not only to quickly obtain the basic connections between knowledge points and build a dynamic graph framework, but also to deeply integrate data from multiple domains, uncover potential connections, optimize weak connections below a threshold, and significantly improve the accuracy of connection strength calculation. Based on the optimized weight distribution, the core paths of the knowledge structure can be accurately located. Through the classification and reorganization of the data set, the fusion effect of data from various domains can be efficiently judged, achieving dynamic optimization of the knowledge graph. This process effectively breaks down cross-industry knowledge barriers, enhances the depth and flexibility of knowledge connections, provides more intelligent and accurate knowledge support for cross-industry information systems, and significantly improves the efficiency and quality of multi-domain data fusion.

[0029] S3. For the correlation strength in the dynamic graph, obtain real-time feedback data on changes in business scenarios. When the scenario changes indicated by the feedback data exceed the preset threshold, trigger the adaptive adjustment mechanism to update the node weights and edge connections in the graph, and obtain the adjusted knowledge structure.

[0030] Furthermore, the adjusted knowledge structure is obtained, specifically including: S31. Through the real-time monitoring system, change feedback data is obtained from the business scenario, the correlation strength in the dynamic graph is continuously tracked, and then a preset threshold is used for comparison and analysis. When the indicator value of the scenario change exceeds the preset threshold, it is determined that the adaptive adjustment mechanism needs to be triggered and the triggering conditions for adjustment are determined. S32. Conduct a preliminary evaluation of the node weights and edge connections in the dynamic graph to identify the weak links in the current knowledge structure, obtain the target regions that need to be optimized, recalculate the node weights and edge connections using a graph neural network model, dynamically adjust the association strength, and obtain the updated graph parameters. S33. Optimize the knowledge structure as a whole by updating the graph parameters, obtain the adjusted node distribution and edge connection status, determine the new graph form, and then verify it with real-time feedback data from business scenarios. S34. When the verification results show that the scene changes are still outside the preset threshold, the adjustment mechanism is repeated to obtain the final stable knowledge structure, record and store the long-term operating status of the dynamic graph, and obtain historical data archives that can be used for subsequent analysis.

[0031] Specifically, by continuously collecting data on changes in business scenarios through a real-time monitoring system and comparing it with preset thresholds, an adaptive adjustment mechanism is precisely triggered, which can quickly locate weak links in the knowledge structure of the dynamic graph. By recalculating node weights, edge connections, and association strengths using a graph neural network model, the knowledge structure is dynamically optimized, and the adjustment effect is ensured through repeated verification until a stable knowledge structure is formed. This process endows the system with real-time perception and proactive adaptation capabilities to changes in business scenarios, breaking the limitations of traditional static knowledge systems, effectively avoiding the disconnect between the knowledge structure and actual needs, and recording historical operation data of the graph to provide data support for subsequent optimization, significantly improving the dynamic adaptability and long-term stability of cross-industry information systems.

[0032] S4. Based on the adjusted knowledge structure and the requirements for trust assurance and information security verification of interactions, a trust verification protocol for cross-domain data interaction is constructed to obtain identity identifiers and data signatures from both parties to determine the integrity and legality of the interactive data.

[0033] Furthermore, determining the completeness and validity of the interactive data specifically includes: By adjusting the knowledge structure, a verification protocol framework is constructed to address the trust verification requirements in cross-domain interactions. The identity identifiers and data signatures are obtained from both parties to determine the initial set of interaction information. The identity identifiers are then verified by comparing them with a pre-established identity database. If the identity identifier does not match the database record, it is marked as an abnormal identity, and the identity verification result is obtained. Further analysis of the data signature is conducted, and a digital signature verification tool is used to decrypt and compare the signature content. If the signature content is inconsistent with the original data, it is marked as a signature anomaly. The signature verification status is then determined, and the data integrity is checked. Key fields are extracted from the data transmitted between the two parties, and a hash algorithm is used to calculate the data fingerprint. If the calculation result does not match the expected fingerprint, it is determined that the data may have been tampered with, and the integrity detection conclusion is obtained. The system assesses data legality by extracting compliance features from the data content and matching them with a pre-defined compliance rule base. If the features do not meet the rule requirements, the data is marked as invalid, and the legality assessment result is determined. The system also performs a comprehensive analysis of interaction security by integrating identity verification, signature status, integrity conclusions, and legality results. If any link is abnormal, a security alarm mechanism is triggered to obtain the final security judgment status. The trust verification protocol for cross-domain interaction is dynamically updated, abnormal data and security alarm information are recorded in the knowledge structure, and the protocol parameters are optimized and adjusted using a graph neural network model to obtain the updated verification framework.

[0034] The verification protocol framework is built upon an adjusted knowledge structure, encompassing a systematic verification process and rule set tailored to the trust verification needs of cross-domain interactions. It involves: collecting identity identifiers and data signatures from both parties to construct an initial set of interaction information; establishing an identity verification mechanism to verify the authenticity of the interacting entities' identities by comparing them with a pre-built identity database; utilizing digital signature verification tools and hash algorithms to decrypt and compare data signatures and verify data integrity; simultaneously, assessing the legality of the data content based on a pre-defined compliance rule base; and finally, integrating the verification results from each stage to form a comprehensive security judgment logic. A graph neural network model is introduced to dynamically optimize protocol parameters based on abnormal data, ensuring the effectiveness and adaptability of the verification process and comprehensively guaranteeing the security and trustworthiness of cross-domain data interactions.

[0035] Specifically, by constructing a cross-domain data interaction trust verification protocol, and combining multiple methods such as identity database comparison, digital signature decryption verification, hash algorithm fingerprint comparison, and compliance rule base matching, the protocol comprehensively detects the identity legitimacy, signature validity, integrity, and content compliance of the interacting data. This enables accurate identification of security risks such as abnormal identities, data tampering, and non-compliant content, triggering real-time alarm mechanisms. Simultaneously, abnormal information is recorded in a knowledge structure, and protocol parameters are dynamically optimized using graph neural networks, forming a closed-loop mechanism of "detection-alarm-optimization." This effectively solves security problems such as identity forgery, data tampering, and insufficient compliance in cross-industry data interaction, constructing a trustworthy cross-domain interaction environment, ensuring the security and reliability of data during cross-industry flow, and improving the overall information security protection capabilities of the system.

[0036] S5. When the integrity and legality verification results of the interactive data pass, the data is allowed to be transmitted in real time within the framework of cross-industry collaborative efficiency, and the time delay and packet loss rate data during the transmission process are obtained to determine the degree to which the real-time interaction requirements are met.

[0037] Furthermore, determine the degree to which real-time interaction requirements are met, specifically including: By analyzing the complete and valid verification results of the interactive data, when the verification result meets the preset standard, the data is triggered to enter the real-time transmission process, obtain the initial transmission status information, and determine whether the transmission is feasible. Using a pre-established transmission channel, real-time transmission operations are performed within a cross-industry business collaboration framework. Time delay data during transmission is collected to obtain delay distribution and packet loss rate information. By comparing with a preset threshold range, it is determined whether the transmission quality meets the real-time standard. When the transmission quality does not meet the real-time standard, the priority strategy of the transmission channel is adjusted, the time delay and packet loss rate data are collected again, the effect of the adjusted transmission is determined, and the support vector machine algorithm is used to comprehensively analyze the time delay and packet loss rate to obtain the evaluation result of the degree of satisfaction of the real-time interaction requirements. By further processing the satisfaction assessment results, if the assessment result is lower than the preset threshold, the backup transmission channel is triggered to reacquire the data indicators during the transmission process and determine the final demand satisfaction status.

[0038] Specifically, it effectively solves the problem of difficulty in ensuring real-time data transmission in cross-industry data transmission, ensures efficient and stable data transmission within the cross-industry collaborative framework, improves the real-time response capability and reliability of data interaction, provides strong support for real-time collaboration of cross-industry businesses, and significantly optimizes the transmission efficiency and user experience of cross-industry information systems.

[0039] S6. Based on the data analysis results of real-time interaction requirements, obtain relevant indicators for resource utilization optimization, extract the distribution of bandwidth usage and computing load during the transmission process, and determine whether there are resource bottlenecks or uneven distribution.

[0040] Furthermore, determining whether there are resource bottlenecks or uneven distribution includes: By analyzing data, bandwidth usage distribution and load distribution are extracted from real-time interactive data to obtain resource usage information. When the bandwidth usage distribution in the resource usage information exceeds a preset threshold, the bandwidth usage distribution is obtained again by dynamically adjusting the transmission channel priority to determine the bottleneck relief status. The load distribution is analyzed and calculated. If there is a concentration of load distribution, the load balancing algorithm is used to redistribute the computing tasks to obtain the adjusted load distribution. Then, by comparing the adjusted load distribution with the preset threshold, it is determined whether there is an uneven distribution. If uneven distribution still exists, a clustering algorithm is used to classify and analyze bandwidth usage and computational load to obtain an optimized resource allocation scheme. The resource scheduling strategy in the transmission process is adjusted to obtain a new distribution of bandwidth usage and computational load. The final resolution of resource bottlenecks and uneven distribution is then determined. Finally, a decision tree algorithm is used to comprehensively evaluate resource optimization indicators to obtain the resource utilization efficiency for real-time interaction requirements.

[0041] Specifically, by extracting bandwidth usage and computational load distribution from real-time interactive data, and combining this with preset thresholds to identify resource bottlenecks and uneven allocation, the system dynamically adjusts transmission channel priorities and uses load balancing algorithms to redistribute tasks. For any remaining uneven allocation, clustering algorithms generate optimization solutions and adjust resource scheduling strategies. Finally, a decision tree algorithm is used to comprehensively evaluate resource utilization efficiency. This approach can accurately pinpoint resource usage pain points in cross-industry data transmission, effectively alleviating issues such as bandwidth overload and concentrated computational load, and achieving intelligent scheduling and optimized allocation of resources. This process significantly improves the system's dynamic response capability to resource bottlenecks, avoids transmission efficiency degradation caused by resource imbalance, and greatly improves resource utilization efficiency and system operational stability in cross-industry collaborative scenarios.

[0042] S7. Based on the judgment results of resource bottlenecks or uneven allocation, adjust the data transmission path and computing task allocation strategy according to the dynamic environment adaptability, obtain the latest resource status information from the system, and determine the optimized resource scheduling scheme.

[0043] Furthermore, the optimized resource scheduling scheme is determined, specifically including: By periodically collecting the status information of system resources, a resource usage distribution map is generated to identify unevenly distributed or bottleneck nodes. When the status information shows that the resource utilization rate of a certain node exceeds a preset threshold, it is marked as a bottleneck node, and a list of bottleneck nodes is obtained. Graph theory algorithms are used to calculate data transmission paths, generate an optimized set of transmission paths, adjust the allocation ratio of computational tasks, generate a task allocation scheme, and use reinforcement learning algorithms to iteratively update the scheduling strategy according to dynamic environmental changes, determine a dynamic scheduling scheme, and adjust system resource allocation in real time to generate the final resource scheduling result.

[0044] Specifically, it effectively solves the problem of resource scheduling lag in dynamic environments, realizes intelligent dynamic optimization of transmission paths and task allocation, enables resource scheduling schemes to evolve in real time with the system resource status, significantly improves the system's adaptability to dynamic environments, avoids resource waste and bottleneck congestion, and ultimately maximizes resource utilization efficiency and system operation efficiency and stability in cross-industry data transmission.

[0045] Furthermore, after determining the optimized resource scheduling scheme, it also includes: updating the priority rules for data interaction and task processing, obtaining monitoring data on execution efficiency and response time from the scheduling logs, and judging the improvement of overall collaborative performance.

[0046] Furthermore, assess the improvement in overall collaborative performance, specifically including: Historical data is obtained from the scheduling logs, and the execution status of resource scheduling and task processing is classified and organized to obtain a preliminary dataset of execution efficiency and response time. Preset thresholds are used to classify the execution efficiency and response time. When the response time of a certain type of task exceeds the threshold, it is marked as a high-priority task. The set of tasks that need to be optimized is determined, the flow order of data interaction is adjusted, and the optimized interaction path is obtained by analyzing the bottleneck links in industry collaboration. Based on the optimized interaction path, the priority rules are reconfigured, the support vector machine algorithm is used to predict the task processing order, the best scheduling scheme for task execution is determined, real-time monitoring data is obtained, and the trend of performance improvement is determined by the fluctuation of execution efficiency and response time. Analyze the changes in cross-industry collaboration efficiency. If the trend does not reach the preset target, iteratively adjust the optimization scheme to obtain a new resource scheduling strategy. Continuously collect monitoring data from the scheduling logs to determine whether the overall collaboration performance meets the expected standards.

[0047] Specifically, it enables dynamic tracking and precise optimization of cross-industry system collaboration performance, effectively solving the problems of fixed priorities and lagging performance improvement in traditional scheduling. It allows the system to proactively optimize task processes based on historical data and real-time monitoring, significantly improving the response speed and execution efficiency of cross-industry collaboration, ensuring that overall performance continues to evolve towards the expected goals, and ultimately achieving high efficiency and stability in system integration.

[0048] Example 2

[0049] Please see Figure 4 This embodiment provides a knowledge graph-based intelligent integration system for cross-industry information systems, used to implement a knowledge graph-based intelligent integration method for cross-industry information systems, including: The data standardization and integration module classifies, unifies the format and semantically annotates multi-source heterogeneous raw data through a cross-industry data mapping model, handles data missing and anomaly issues, extracts features and fuses data using the support vector machine algorithm, and finally forms a standardized preliminary integrated dataset, laying the foundation for subsequent processing. The knowledge graph construction and optimization module, based on integrated data, uses graph neural networks to construct dynamic knowledge graphs, calculates the association strength and weight distribution of knowledge points, integrates data from multiple domains to mine potential connections, optimizes graph edge relationships, clarifies core connection paths, and realizes dynamic adjustment of knowledge structure and deep integration of data from multiple domains. The adaptive adjustment module monitors the feedback of changes in business scenarios in real time. When the scenario indicators exceed the threshold, the adaptive mechanism is triggered to re-evaluate and calculate the weights and edge connections of the graph nodes, optimize the knowledge structure and verify the effect, and continue to iterate until it is stable. At the same time, the historical operation data of the graph is recorded to support subsequent analysis. The cross-domain trust verification module constructs a cross-domain interaction trust verification protocol, verifies the identities and data signatures of both parties involved in the interaction, detects data integrity and legality through hash algorithms and compliance rule bases, integrates multi-dimensional verification results to trigger security alarms, and dynamically updates protocol parameters to ensure the trustworthiness and security of data interaction. The real-time transmission and performance evaluation module, after the data verification is passed, realizes cross-industry real-time data transmission through a preset channel, collects indicators such as latency and packet loss rate to evaluate transmission quality, and adjusts the transmission strategy or activates the backup channel when the standard is not met. The algorithm comprehensively analyzes the data to determine the degree to which the real-time requirements are met. The resource scheduling optimization module analyzes bandwidth usage and computational load data during transmission, identifies resource bottlenecks and uneven allocation issues, optimizes resource scheduling by adjusting transmission paths and load balancing algorithms, generates dynamic scheduling schemes by combining graph theory and reinforcement learning algorithms, and iteratively optimizes collaborative performance based on log data to improve resource utilization efficiency.

[0050] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A cross-industry information system intelligent integration method based on knowledge graphs, characterized in that: Includes the following steps: By establishing a cross-industry data mapping model, the heterogeneous characteristics of the data are standardized. The original datasets are obtained from data sources in multiple industries, and the fields are aligned and semantically labeled according to a unified format to obtain a preliminary integrated data set. Graph neural networks are used to model knowledge structure adjustment. To meet the needs of knowledge association depth and multi-domain data fusion, a dynamic graph of node and edge relationships is constructed to determine the association strength and weight distribution between knowledge points. Obtain real-time feedback data on changes in business scenarios, determine whether the scenario changes indicated by the feedback data exceed a preset threshold, and if so, trigger an adaptive adjustment mechanism to update the node weights and edge connections in the graph to obtain the adjusted knowledge structure. To address the need for trusted interaction assurance and information security verification, a trust verification protocol for cross-domain data interaction is constructed to obtain identity identifiers and data signatures from both parties involved in the interaction, and to determine the integrity and legality of the interacting data. When the integrity and legality verification of the interactive data pass, the data is allowed to be transmitted in real time within the framework of cross-industry collaborative efficiency. The time delay and packet loss rate data during the transmission process are obtained to determine the degree to which the real-time interaction requirements are met. Obtain relevant indicators for resource utilization optimization, extract the distribution of bandwidth usage and computing load during the transmission process, and determine whether there are resource bottlenecks or uneven distribution phenomena. Based on the judgment of resource bottlenecks or uneven distribution, the system adapts to the dynamic environment, adjusts the data transmission path and computing task allocation strategy, obtains the latest resource status information, and determines the optimized resource scheduling scheme.

2. The method for intelligent integration of cross-industry information systems based on knowledge graphs according to claim 1, characterized in that: The resulting preliminarily integrated dataset includes: The original dataset is obtained by cross-industry data sources, and a cross-industry data mapping model is built to perform preliminary classification of the data. The heterogeneous characteristics are standardized. When the data field formats are inconsistent, the format is unified by field alignment tools to determine the dataset with consistent format. For semantic annotation requirements, the data is annotated using preset semantic recognition rules. In combination with industry characteristics and cross-industry requirements, when there is missing or abnormal data, the data completion mechanism is used to determine the complete dataset. The support vector machine algorithm is used to extract features from the data, and data fusion technology is used to integrate the data set construction requirements. Then, semantic disambiguation rules are used to adjust the data with semantic conflicts, resulting in an optimized data set.

3. The method for intelligent integration of cross-industry information systems based on knowledge graphs according to claim 1, characterized in that: Determining the strength of associations and weight distribution between knowledge points specifically includes: By integrating the dataset, a graph neural network is used to initially model the knowledge structure, obtain the basic connection relationships between knowledge points, determine the initial node relationships, construct a dynamic graph of edge relationships, calculate the initial association strength between each node, and obtain the structural framework of the dynamic graph. By integrating data from multiple fields and adjusting the deep calculation method of knowledge association, the potential connection strength between each knowledge point is determined. When the potential connection strength is lower than the preset threshold, the edge relationship construction is optimized and adjusted, the association strength is recalculated, and a more accurate weight distribution is obtained. Based on the optimized weight distribution, the node and edge relationships in the dynamic graph are updated to determine the core connection paths in the knowledge structure. Then, the matching degree of knowledge association and multi-domain integration is analyzed, the data set is classified and reorganized, the integration effect of data in each domain is judged, and the final dynamic adjustment result is determined.

4. The method for intelligent integration of cross-industry information systems based on knowledge graphs according to claim 1, characterized in that: The adjusted knowledge structure includes: By using a real-time monitoring system, change feedback data is obtained from business scenarios, the correlation strength in the dynamic graph is continuously tracked, and then a comparison analysis is performed using preset thresholds. When the indicator value of the scenario change exceeds the preset threshold, it is determined that an adaptive adjustment mechanism needs to be triggered, and the triggering conditions for adjustment are determined. A preliminary assessment of node weights and edge connections in the dynamic graph is conducted to identify weak links in the current knowledge structure and determine the target regions that need optimization. A graph neural network model is then used to recalculate node weights and edge connections, dynamically adjust the association strength, and obtain updated graph parameters. By updating the graph parameters, the knowledge structure is optimized as a whole, the adjusted node distribution and edge connection status are obtained, the new graph form is determined, and then the results are verified using real-time feedback data from business scenarios. If the verification results show that the scene changes are still outside the preset threshold, the adjustment mechanism is repeated to obtain the final stable knowledge structure and record and store the long-term operating status of the dynamic graph.

5. The method for intelligent integration of cross-industry information systems based on knowledge graphs according to claim 1, characterized in that: Determining the integrity and validity of interactive data specifically includes: By adjusting the knowledge structure, a verification protocol framework is constructed to address the trust verification requirements in cross-domain interactions. The identity identifiers and data signatures are obtained from both parties involved in the interaction. The identity identifiers are then verified and compared with a pre-established identity database to obtain the identity verification results. Further analysis of the data signature is conducted, and the signature content is decrypted and compared using a digital signature verification tool. The signature verification status is then determined, and the data integrity is checked. Key fields are extracted from the data transmitted between the two parties, and a hash algorithm is used to calculate the data fingerprint to obtain the integrity check conclusion. The system assesses data legality, extracts compliance features from data content, matches them with a pre-defined compliance rule base, determines the legality assessment result, conducts a comprehensive analysis of interaction security, integrates identity verification, signature status, integrity conclusions and legality results, and triggers a security alarm mechanism when any link is abnormal to obtain the final security judgment status. The trust verification protocol for cross-domain interaction is dynamically updated, abnormal data and security alarm information are recorded in the knowledge structure, and the protocol parameters are optimized and adjusted using a graph neural network model to obtain the updated verification framework.

6. The method for intelligent integration of cross-industry information systems based on knowledge graphs according to claim 1, characterized in that: Determine the degree to which real-time interaction requirements are met, specifically including: By analyzing the complete and valid verification results of the interactive data, when the verification result meets the preset standard, the data is triggered to enter the real-time transmission process and obtain the initial transmission status information. Using a pre-established transmission channel, real-time transmission operations are performed within a cross-industry business collaboration framework. Time delay data is collected during the transmission process, and packet loss rate information is obtained. By comparing with a preset threshold range, it is determined whether the transmission quality meets the real-time standard. When the transmission quality does not meet the real-time standard, the priority strategy of the transmission channel is adjusted, the time delay and packet loss rate data are collected again, and the support vector machine algorithm is used to comprehensively analyze the time delay and packet loss rate to obtain the evaluation result of the degree of satisfaction of the real-time interaction requirements. By further processing the satisfaction assessment results, if the assessment result is lower than the preset threshold, the backup transmission channel is triggered to reacquire the data indicators during the transmission process and determine the final demand satisfaction status.

7. The method for intelligent integration of cross-industry information systems based on knowledge graphs according to claim 1, characterized in that: To determine whether there are resource bottlenecks or uneven distribution, the following should be considered: By analyzing data, bandwidth usage distribution and load distribution are extracted from real-time interactive data to obtain resource usage information. When the bandwidth usage distribution in the resource usage information exceeds a preset threshold, the bandwidth usage distribution is obtained again by dynamically adjusting the transmission channel priority to determine the bottleneck relief status. The load distribution is analyzed and calculated. If there is a concentration of load distribution, the load balancing algorithm is used to redistribute the computing tasks to obtain the adjusted load distribution. Then, by comparing the adjusted load distribution with the preset threshold, it is determined whether there is an uneven distribution. If uneven distribution still exists, a clustering algorithm is used to classify and analyze bandwidth usage and computational load, adjust the resource scheduling strategy of the transmission process, obtain new bandwidth usage and computational load distribution, determine the final resolution of resource bottlenecks and uneven distribution, and then use a decision tree algorithm to comprehensively evaluate resource optimization indicators to obtain the resource utilization efficiency for real-time interaction requirements.

8. The method for intelligent integration of cross-industry information systems based on knowledge graphs according to claim 1, characterized in that: The optimized resource scheduling scheme is determined, specifically including: By periodically collecting the status information of system resources, a resource usage distribution map is generated to identify unevenly distributed or bottleneck nodes. When the status information shows that the resource utilization rate of a certain node exceeds a preset threshold, it is marked as a bottleneck node, and a list of bottleneck nodes is obtained. Graph theory algorithms are used to calculate data transmission paths, generate an optimized set of transmission paths, adjust the allocation ratio of computational tasks, generate a task allocation scheme, and use reinforcement learning algorithms to iteratively update the scheduling strategy according to dynamic environmental changes, determine a dynamic scheduling scheme, and adjust system resource allocation in real time to generate the final resource scheduling result.

9. The method for intelligent integration of cross-industry information systems based on knowledge graphs according to claim 8, characterized in that: After determining the optimized resource scheduling scheme, the process also includes: updating the priority rules for data interaction and task processing, obtaining monitoring data on execution efficiency and response time from the scheduling logs, and judging the improvement in overall collaborative performance.

10. A knowledge graph-based intelligent integration system for cross-industry information systems, used to implement the knowledge graph-based intelligent integration method for cross-industry information systems as described in any one of claims 1-9, characterized in that: include: The data standardization and integration module classifies, unifies the format and semantically labels multi-source heterogeneous raw data through a cross-industry data mapping model, extracts features and fuses data using the support vector machine algorithm, and finally forms a standardized preliminary integrated dataset. The knowledge graph construction and optimization module, based on integrated data, constructs a dynamic knowledge graph through graph neural networks, calculates the association strength and weight distribution of knowledge points, integrates data from multiple domains to mine potential connections, optimizes graph edge relationships, clarifies core connection paths, and performs dynamic adjustments to the knowledge structure and deep integration of data from multiple domains. The adaptive adjustment module monitors the feedback of changes in business scenarios in real time. When the scenario indicators exceed the threshold, the adaptive mechanism is triggered to re-evaluate and calculate the weights and edge connections of the graph nodes, optimize the knowledge structure and verify the effect, and continue to iterate until it is stable. The cross-domain trust verification module constructs a cross-domain interaction trust verification protocol, verifies the identities and data signatures of both parties involved in the interaction, detects data integrity and legality through hash algorithms and compliance rule bases, integrates multi-dimensional verification results to trigger security alarms, and dynamically updates protocol parameters. The real-time transmission and performance evaluation module, after the data verification is passed, realizes cross-industry real-time data transmission through a preset channel, collects latency and packet loss rate indicators to evaluate transmission quality, and adjusts the transmission strategy or activates the backup channel when the standard is not met. The algorithm comprehensively analyzes the data to determine the degree to which the real-time requirements are met. The resource scheduling optimization module analyzes bandwidth usage and computational load data during transmission, generates dynamic scheduling schemes by optimizing resource scheduling in combination with graph theory and reinforcement learning algorithms, and iteratively optimizes collaborative performance based on log data.

Images