Data fusion and knowledge modeling system and method

Abstract

The invention provides a data fusion and knowledge modeling system and method, and the system comprises a data collection module which is used for collecting multi-source heterogeneous data in real time according to a collection demand; the data preprocessing module is used for carrying out self-adaptive preprocessing on the multi-source heterogeneous data to obtain processed standard data; the visual mapping module is used for performing visual mapping on the standard data according to the data characteristics and the user requirements to obtain visual data; the data fusion module is used for intelligently fusing the visual data; the dynamic ontology construction module is used for dynamically constructing an ontology according to the fused data; and the front-end visual interaction module is used for providing a front-end visual interface and a multi-type interaction interface for a user. Through an innovative data processing architecture and a visual interaction mechanism, the problem of multi-source heterogeneous data fusion is effectively solved, dynamic semantic modeling of data is realized, and a visual and efficient data processing and knowledge discovery tool is provided for a user.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a data fusion and knowledge modeling system and method. BACKGROUND

[0002] With the digital wave sweeping across various industries, multi-source heterogeneous data is growing explosively, and the access barriers caused by data silos, the processing complexity caused by the non-uniform formats, and the understanding deviation caused by semantic ambiguity have brought unprecedented technical challenges to the intelligent fusion and deep analysis of data. Traditional data fusion methods usually require a large amount of manual data cleaning, conversion and mapping work, which is low in efficiency and prone to errors, and is difficult to cope with the ambiguity and dynamic changes of data semantics, and is prone to information loss or incorrect association in the fusion process, which cannot meet the needs of deep data analysis and knowledge mining.

[0003] In the aspect of ontology construction, the existing ontology construction technology is usually a static construction method. When the data is updated, the constructed ontology structure cannot be timely perceived and adaptively adjusted, resulting in disconnection between the ontology structure and the actual data semantics, and the inability to accurately reflect the latest relationship between data, which seriously restricts the effective expression and reasoning application of knowledge. In the aspect of visual interaction, the existing data fusion process is mostly in the form of a black box, lacking intuitive visual display and interactive means, and users cannot understand the logic of data fusion, nor can they flexibly adjust the analysis process according to their own needs, greatly affecting the efficiency and quality of data processing. SUMMARY

[0004] The purpose of the present application is to provide a data fusion and knowledge modeling system and method, which aims to solve the above problems in the prior art.

[0005] The data fusion and knowledge modeling system provided by the embodiments of the present application comprises: A data acquisition module connected with a data preprocessing module and a front-end visual interaction module, configured to acquire multi-source heterogeneous data in real time according to acquisition requirements, and transmit the multi-source heterogeneous data to the data preprocessing module; A data preprocessing module connected with the data acquisition module and a visual mapping module, configured to receive the multi-source heterogeneous data, adaptively preprocess the multi-source heterogeneous data to obtain standard data, and transmit the standard data to the visual mapping module; A visual mapping module connected with the data preprocessing module, a data fusion module and a front-end visual interaction module, configured to receive the standard data, and perform visual mapping on the standard data according to data characteristics and user requirements to obtain visual data, and send the visual data to the data fusion module; The data fusion module is connected with the visualization mapping module, the dynamic ontology construction module and the front-end visualization interaction module, and is configured to receive the visualization data, intelligently fuse the visualization data, and transmit the fused data to the dynamic ontology construction module. The dynamic ontology construction module is connected with the data fusion module and the front-end visualization interaction module, and is configured to receive the fused data, and dynamically construct an ontology according to the fused data by using an incremental learning algorithm and a consistency checking mechanism. The front-end visualization interaction module is connected with the data acquisition module, the visualization mapping module, the data fusion module and the dynamic ontology construction module, and is configured to provide a front-end visualization interface and a multi-type interaction interface for a user.

[0006] The embodiment of the present application provides a data fusion and knowledge modeling method, which comprises the following steps: The data acquisition module acquires multi-source heterogeneous data in real time according to acquisition requirements, and transmits the multi-source heterogeneous data to a data preprocessing module. The data preprocessing module receives the multi-source heterogeneous data, adaptively preprocesses the multi-source heterogeneous data, obtains standard data after processing, and transmits the standard data to a visualization mapping module. The visualization mapping module receives the standard data, and performs visualization mapping on the standard data according to data characteristics and user requirements, to obtain visualization data, which is sent to a data fusion module. The data fusion module receives the visualization data, intelligently fuses the visualization data, and transmits fused data to a dynamic ontology construction module. The dynamic ontology construction module receives the fused data, and dynamically constructs an ontology according to the fused data by using an incremental learning algorithm and a consistency checking mechanism. The front-end visualization interaction module provides a front-end visualization interface and a multi-type interaction interface for a user.

[0007] The embodiment of the present application also provides an electronic device, which comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the computer program is executed by the processor to implement the steps of the data fusion and knowledge modeling method.

[0008] The embodiment of the present application also provides a computer readable storage medium, which stores an implementation program of information transmission, and the program is executed by a processor to implement the steps of the data fusion and knowledge modeling method.

[0009] The embodiment of the present application can have the following beneficial effects: the embodiment of the present application aims to provide a data fusion and knowledge modeling system and method based on front-end visualization, effectively solves the problem of multi-source heterogeneous data fusion through an innovative data processing architecture and a visual interactive mechanism, realizes dynamic semantic modeling of data, provides users with intuitive and efficient data processing and knowledge discovery tools, and meets the urgent needs of various industries for deep analysis and intelligent decision-making of complex data. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the one or more embodiments of the present specification or the prior art, the drawings needed in the embodiment or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments described in the present specification, and other drawings can be obtained by those skilled in the art without creative labor.

[0011] Figure 1 is a data fusion and knowledge modeling system schematic diagram of the embodiment of the present application; Figure 2 is a data fusion and knowledge modeling method flow chart of the embodiment of the present application. DETAILED DESCRIPTION

[0012] In order to make the person skilled in the art better understand the technical solutions in the one or more embodiments of the present specification, the technical solutions in the one or more embodiments of the present specification will be clearly and completely described below in combination with the drawings in the one or more embodiments of the present specification. Obviously, the described embodiments are only some embodiments of the present specification, not all embodiments. Based on the one or more embodiments of the present specification, all other embodiments obtained by those skilled in the art without creative labor should belong to the protection scope of the present document.

[0013] System embodiment According to the embodiment of the present application, a data fusion and knowledge modeling system is provided, Figure 1 is a data fusion and knowledge modeling system schematic diagram of the embodiment of the present application, as Figure 1 shown, the data fusion and knowledge modeling system according to the embodiment of the present application specifically includes: The data acquisition module 10 is connected with the data preprocessing module and the front-end visual interactive module, and is used for acquiring multi-source heterogeneous data in real time according to the acquisition demand, and transmitting the multi-source heterogeneous data to the data preprocessing module; The data acquisition module supports a plurality of data acquisition protocols and interfaces, and specifically includes: The data source scanning and identifying unit is connected with the protocol adaptation and data collecting unit, and is configured to automatically identify different types of data sources through protocol feature analysis, and complete corresponding authentication or configuration; The protocol adaptation and data collecting unit is connected with the data source scanning and identifying unit, and is configured to match corresponding data collecting protocols according to the identified data source types, collect multi-source heterogeneous data according to preset frequencies or trigger conditions, and start system retry or notify corresponding managers in the case of connection timeout.

[0014] The data preprocessing module 11 is connected with the data collecting module and the visual mapping module, and is configured to receive the multi-source heterogeneous data, adaptively preprocess the multi-source heterogeneous data, obtain processed standard data, and transmit the standard data to the visual mapping module, and is specifically configured to: repeat data is deleted, and unstructured data is compared through MD5 hash values, and the latest record in the data with the same hash value is retained; The missing proportion of numerical value type data is counted, if the missing proportion is less than or equal to a preset proportion, the mean value of the corresponding field is used to fill the missing value, if the missing proportion is greater than the preset proportion, a random forest algorithm is used to construct a missing value prediction model, and the missing value is predicted and filled through the missing value prediction model, and text type missing values are uniformly filled; Based on the principle of 3 The abnormal value of numerical value type data is detected, and the mean value of the adjacent normal value is used to replace the detected abnormal value, the isolated forest algorithm is used to detect the abnormal value of time series data, and the detected abnormal value is corrected through regression analysis or interpolation method; The processed multi-source heterogeneous data is subjected to data format unification, data type conversion and normalization processing.

[0015] The visual mapping module 12 is connected with the data preprocessing module, the data fusion module and the front-end visual interactive module, and is configured to receive the standard data, and perform visual mapping on the standard data according to data characteristics and user demands, obtain visual data, and send the visual data to the data fusion module, and specifically includes: The feature analysis unit is connected with the visual template adjustment unit, and is configured to extract data characteristics of the standard data, and recommend a preliminary visual template for a user based on the data characteristics and user historical behaviors through a recommendation algorithm; The visual template adjustment unit is connected with the feature analysis unit and the data field binding unit, and is configured to receive user demands, determine a used preliminary visual template according to the user demands, and support user self-defined adjustment of the used preliminary visual template. The data field binding unit is connected to the visualization template adjustment unit. It is used to receive the adjusted visualization template information and bind the data fields to the attributes of the visualization elements in the visualization template information to complete the conversion of data into visualization graphics.

[0016] The data fusion module 13, connected to the visualization mapping module, the dynamic ontology construction module, and the front-end visualization interaction module, is used to receive the visualization data, intelligently fuse the visualization data, and transmit the fused data to the dynamic ontology construction module. Specifically, it includes: A semantic feature extraction unit, connected to a semantic similarity calculation unit, is used to perform adaptive feature extraction on the visualized data to obtain a unified semantic representation result; A semantic similarity calculation unit, connected to the semantic feature extraction unit and the machine learning fusion unit, is used to calculate the semantic similarity between different data based on the semantic representation results using a cosine similarity algorithm. The machine learning fusion unit, connected to the semantic similarity calculation unit, is used to merge similar data using the DBSCAN density clustering algorithm based on the semantic similarity, and to classify and disambiguate the fused data using the decision tree algorithm.

[0017] The dynamic ontology construction module 14, connected to the data fusion module and the front-end visualization interaction module, is used to receive the fused data and dynamically construct an ontology based on the fused data using an incremental learning algorithm and a consistency verification mechanism. Specifically, it includes: An initial ontology construction unit, connected to an incremental learning unit, is used to construct an initial ontology based on the fused data using the OWL language, and to define the classes, attributes, and relationships of the initial ontology; An incremental learning unit, connected to the initial ontology construction unit and the consistency verification unit, is used to monitor data in real time. When the data changes, the Jena framework and ontology mapping algorithm are used to perform semantic matching between the new data and the initial ontology, and the ontology structure is updated or expanded according to the matching results. The consistency verification unit, connected to the incremental learning unit, is used to perform consistency checks on the updated or expanded ontology using the Pellet inference engine, and to automatically roll back or adjust the ontology structure when conflicts are found.

[0018] The front-end visualization interaction module 15 is connected to the data acquisition module, visualization mapping module, data fusion module and dynamic ontology construction module, and is used to provide users with a front-end visualization interface and multiple types of interactive interfaces. The front-end visualization interaction module comprises a visualization result rendering unit, an interactive operation processing unit and a personalized customization unit which are in communication with each other. The visualization result rendering unit is configured to render the data fusion process, the ontology structure and the data relationship into a front-end visualization interface and support adaptive zooming of the front-end visualization interface. The interactive operation processing unit is configured to receive user operation information and complete multi-type interaction between the user and the system according to the operation information. The personalized customization unit is configured to support user-defined visualization effect parameters and save the customized visualization effect as a template to the system.

[0019] The above technical solutions of the embodiments of the present application will be described in detail below in combination with the specific conditions of the data fusion and knowledge modeling system of the embodiments of the present application.

[0020] The embodiments of the present application provide a data fusion and knowledge modeling system based on front-end visualization, and the specific operation contents are as follows. Step 1: Data acquisition The module is configured to acquire heterogeneous data from multiple different data sources, and the data sources include but are not limited to databases, file systems, sensor networks and the like. The module supports multiple data acquisition protocols and interfaces and can adapt to the data acquisition requirements of different types of data sources.

[0021] 1. Data source scanning and identification: after the system is started, a port scanning program is started to detect the network environment and local storage. Through protocol feature analysis, relational databases such as MySQL and Oracle, file systems such as Hadoop distributed file system or cloud storage, and sensor network data sources based on ZigBee and LoRa protocols are automatically identified. For databases, the user needs to manually input the IP address, port number, username and password to complete authentication; the file system specifies the access path; the sensor network configures the communication protocol parameters and node ID.

[0022] 2. Protocol adaptation and data acquisition: according to the type of the identified data source, the acquisition protocols such as HTTP, FTP, Modbus and JDBC are automatically matched. After the connection is established, for periodically changing data, the data is collected at a preset frequency; for non-periodic data, the collection task is started according to the trigger condition. If the connection times out, the system automatically retries 3 times, and if it still fails, the error information is recorded to a log file, and the administrator is notified by email or system message.

[0023] Step 2: Data preprocessing The collected heterogeneous data is cleaned, converted and normalized. Data cleaning includes removing duplicate data, handling missing values and outliers; converting data of different formats to a unified data format; normalization processing makes data have the same measurement standard, facilitating subsequent data fusion operations.

[0024] 1. Duplicate data processing: For structured data, duplicate records are deleted directly from the dataset by comparing the primary key fields of the database table; for unstructured data, fingerprints are generated by MD5 hash value comparison, and the latest record in the data with the same hash value is retained.

[0025] 2. Missing value processing: Calculate the missing proportion of numerical data, if the missing proportion is less than 10%, use the mean value of the field to fill; if it is higher than 10%, use the random forest algorithm to construct a prediction model, and use other related field data to predict and fill the missing values. For text data, all missing values are filled with the "unknown" identifier.

[0026] 3. Outlier processing: For numerical data, based on the 3σ principle, calculate the data mean and standard deviation, and replace the data outside the mean ± 3 times the standard deviation with the mean of the adjacent normal values; for time series data, use the Isolation Forest algorithm to detect outliers, and correct outliers by regression analysis or interpolation method.

[0027] 4. Format unification: Convert CSV, JSON, XML and other structured data to standard JSON format through data parser to ensure data structure consistency; for unstructured data, use OpenCV library to extract 256-dimensional feature vectors for vector representation, and use Librosa library to convert audio data to mel spectrum graph data, achieving structured conversion of unstructured data.

[0028] 5. Type conversion: Convert string type date data to standard timestamp format according to ISO 8601 standard; convert Boolean value data to numerical form with 0 (false) and 1 (true).

[0029] 6. Normalization processing: For numerical data, use the min-max normalization method to map data to the [0, 1] interval, and through normalization processing, eliminate the dimensional differences between different data dimensions, facilitating subsequent data fusion operations.

[0030] Step 3: Visual mapping According to the characteristics of the data and user needs, the preprocessed data is mapped to visual elements such as charts, graphs, animations, etc. This module provides multiple visualization templates and custom visualization configuration options, and users can set up visual mapping through the front-end interface.

[0031] 1. Data feature analysis: Use the data analysis engine to extract the dimension information and distribution characteristics of the data, combine historical user operation data, and preliminarily filter suitable visualization templates through recommendation algorithms, such as column charts for data comparison, line charts for trend analysis, and heat maps for correlation display.

[0032] 2. Visualization mapping settings: Users can select appropriate visualization templates from the recommended template list on the front-end visualization interface and adjust parameters such as element color, size, and position. For custom visualization needs, the system provides a graphical editing interface that supports users to add, delete, and modify visualization elements and set animation effects for elements.

[0033] 3. Data mapping execution: According to user settings on the front-end interface, bind data fields with the properties of visualization elements. For example, map numerical fields to the height of column charts and the Y-axis coordinates of line charts; map categorical fields to column chart category labels and pie chart sector identifiers, thereby completing the conversion from data to visualization graphics.

[0034] Step 4: Data fusion Based on semantic similarity calculation and machine learning algorithms, intelligently fuse the data after visualization mapping. By establishing semantic associations between data, eliminate ambiguity and redundancy, and effectively fuse multi-source heterogeneous data.

[0035] 1. Semantic feature extraction: For text data, use the BERT pre-trained language model to extract 768-dimensional semantic vectors to capture deep semantic information of the text; for numerical and structured data, use the Apriori association rule mining algorithm with a minimum support of 0.1 and a minimum confidence of 0.7 to mine the implicit semantic features and association relationships between data.

[0036] 2. Semantic similarity calculation: Calculate the semantic similarity between data using the cosine similarity algorithm. When the similarity between data is higher than the preset threshold, it is considered as semantically related data.

[0037] 3. Machine learning fusion: Based on the extracted semantic features, use the DBSCAN density clustering algorithm to perform clustering analysis, set the neighborhood radius ε to 0.5 and the minimum sample size MinPts to 5, and merge data of the same type; for data with semantic ambiguity, use the decision tree algorithm to build a classification model, learn classification rules through training data, classify ambiguous data, eliminate data ambiguity, and effectively fuse multi-source heterogeneous data.

[0038] Step 5: Dynamic ontology construction According to the fused data, an incremental learning algorithm is used to dynamically construct the ontology. This module can monitor the changes in data in real time, and automatically update the ontology structure and semantic relationships when new data is added or existing data changes, maintaining the dynamic nature and accuracy of the ontology.

[0039] 1. Initial ontology construction: Based on the fused data, an initial ontology model is constructed using OWL. Classes, properties, and relationships are defined, such as creating a "sensor data" class and adding "temperature," "humidity," and other properties. The "device" class and the "sensor data" class are established with a "generate" relationship, and the ontology structure is designed and defined through the ontology editing tool.

[0040] 2. Incremental learning update: The system monitors data changes in real time, and when new data is added or existing data is modified, the update process is started. Using the Jena framework, the semantics of new data are matched with the existing ontology through the ontology mapping algorithm, and the matching degree is calculated. If the matching degree is higher than 0.7, the corresponding class or property information is updated; if the matching degree is lower than 0.7, a new class or relationship is created to extend and update the ontology structure.

[0041] 3. Consistency check: After updating the ontology, the Pellet inference engine is used to check the consistency of the ontology model. The inheritance relationship between classes, the value range of properties, and other contradictions or conflicts are checked. If a contradiction is found, the system automatically backtracks the update operation, adjusts the ontology structure, and ensures the accuracy and logical consistency of the ontology.

[0042] Step 6: Front-end visual interaction A user-friendly visual interface is provided for users to view the data fusion process, ontology structure, and data relationships, and to perform data filtering, querying, analysis, and other operations. At the same time, users can adjust and customize the visual effects to meet individual needs.

[0043] 1. Interface rendering and display: In the browser environment, AntV, Echarts, and other front-end technologies are used to render the visual interface and display the data fusion result graph and the graph data of the dynamic ontology structure. Various complex network graphs can be drawn and interacted with, showing complex relationship data such as knowledge graphs and social networks, as well as adaptive zooming function that automatically adjusts the chart layout and element size when the browser window size changes, maintaining good visual effects.

[0044] 2. Implementation of interactive operations: Users interact with the visual interface through mouse clicks, drags, and other operations. On the graph, users can select and click on part of the data to view detailed information, and input query conditions to filter and query data. For the dynamic ontology structure graph, users can expand / collapse class nodes to view the relationship between nodes, property definitions, and other detailed content.

[0045] 3、Visual customization adjustment: users can customize the visual effect in the front-end interface, including modifying the pattern style, adding annotation information, adjusting the graphic layout, etc. The customized visual effect can be saved as a template for direct calling next time, meeting the individual needs of different users.

[0046] Method embodiment According to the embodiment of the present application, a data fusion and knowledge modeling method is provided, Figure 2 is the flow chart of the data fusion and knowledge modeling method of the embodiment of the present application, as Figure 2 shown, the data fusion and knowledge modeling method according to the embodiment of the present application specifically includes: Step S201, collecting multi-source heterogeneous data in real time according to the collection requirements through the data collection module, and transmitting the multi-source heterogeneous data to the data preprocessing module; Step S202, receiving the multi-source heterogeneous data through the data preprocessing module, adaptively preprocessing the multi-source heterogeneous data to obtain the processed standard data, and transmitting the standard data to the visualization mapping module; Step S203, receiving the standard data through the visualization mapping module, and visualizing the standard data according to the data characteristics and user requirements to obtain the visualization data, and sending the visualization data to the data fusion module; Step S204, receiving the visualization data through the data fusion module, intelligently fusing the visualization data, and transmitting the fused data to the dynamic ontology construction module; Step S205, receiving the fused data through the dynamic ontology construction module, and dynamically constructing the ontology according to the fused data using the incremental learning algorithm and the consistency checking mechanism; Step S206, providing the front-end visualization interface and the multi-type interaction interface for the user through the front-end visualization interaction module.

[0047] The embodiment of the present application is a method embodiment corresponding to the above-mentioned system embodiment, and the specific operation of each step can be understood with reference to the description of the system embodiment, which will not be repeated here.

[0048] In summary, the embodiment of the present application has the following beneficial effects: 1. Accuracy improvement: through semantic similarity calculation and multiple machine learning algorithms to deeply mine data semantic association, compared with traditional artificial rule conversion, data fusion can be more accurately realized, and information loss and error association are reduced. 2. Enhanced dynamic adaptability: the dynamic ontology construction module can sense data changes in real time and automatically update the ontology, which changes the defect that the existing static ontology construction cannot adapt to dynamic data update. 3. Visualized interactive innovation: Constructing a visualized data space and rich interactive components, intuitively presenting data fusion and ontology construction process, users can understand data relationships through simple interactive operations, which is not available in existing black-box data fusion technology.

[0049] 4. Modularization and customization: The system adopts modular design, and the interfaces of each module are standardized for easy expansion; at the same time, it supports user-defined visual mapping and interactive settings to meet individual needs. 5. Improve decision-making efficiency: Intuitive visual interaction enables non-professional users to quickly understand data, promotes data-driven decision-making process, and improves decision-making efficiency and accuracy. 6. Reduce processing cost: Automated data processing reduces human intervention, reduces data processing cost, and improves data processing efficiency. 7. Expand application scenarios: Strong versatility and expandability make the system widely applicable in medical, financial, transportation and other fields, meet the complex data processing needs of different industries, and have broad application prospects and commercial value.

[0050] Device embodiment one The embodiment of the application provides an electronic device, comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the computer program is executed by the processor to realize the steps as described in the method embodiment.

[0051] Device embodiment two The embodiment of the application provides a computer readable storage medium, wherein the computer readable storage medium stores an implementation program of information transmission, and the program is executed by a processor to realize the steps as described in the method embodiment.

[0052] The computer readable storage medium described in the embodiment includes but is not limited to ROM, RAM, magnetic disk or optical disk, etc.

[0053] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the application, but not to limit them; although the application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the application.

Claims

1. A data fusion and knowledge modeling system, characterized in that... include: The data acquisition module, connected to the data preprocessing module and the front-end visualization interaction module, is used to acquire multi-source heterogeneous data in real time according to the acquisition requirements, and transmit the multi-source heterogeneous data to the data preprocessing module. A data preprocessing module, connected to the data acquisition module and the visualization mapping module, is used to receive the multi-source heterogeneous data, perform adaptive preprocessing on the multi-source heterogeneous data to obtain processed standard data, and transmit the standard data to the visualization mapping module. The visualization mapping module is connected to the data preprocessing module, the data fusion module, and the front-end visualization interaction module. It is used to receive the standard data, perform visualization mapping on the standard data according to data characteristics and user needs to obtain visualization data, and send the visualization data to the data fusion module. The data fusion module is connected to the visualization mapping module, the dynamic ontology construction module and the front-end visualization interaction module. It is used to receive the visualization data, intelligently fuse the visualization data, and transmit the fused data to the dynamic ontology construction module. The dynamic ontology construction module is connected to the data fusion module and the front-end visualization interaction module. It is used to receive the fused data and dynamically construct the ontology based on the fused data using an incremental learning algorithm and a consistency verification mechanism. The front-end visualization interaction module is connected to the data acquisition module, visualization mapping module, data fusion module, and dynamic ontology construction module, and is used to provide users with a front-end visualization interface and multiple types of interactive interfaces.

2. The system according to claim 1, characterized in that, The data acquisition module supports several data acquisition protocols and interfaces, specifically including: The data source scanning and identification unit, connected to the protocol adaptation and data acquisition unit, is used to automatically identify different types of data sources through protocol feature analysis and complete the corresponding authentication or configuration. The protocol adapter and data acquisition unit is connected to the data source scanning and identification unit. It is used to match the corresponding data acquisition protocol according to the identified data source type, and to collect multi-source heterogeneous data according to a preset frequency or triggering conditions. In the event of a connection timeout, it will initiate a system retry or notify the relevant management personnel.

3. The system according to claim 1, characterized in that, The data preprocessing module is specifically used for: The system performs deduplication checks on structured data based on the primary key field, deleting duplicate data; it also compares unstructured data using MD5 hash values, retaining the latest record among data with the same hash value. The missing value ratio of numerical data is statistically analyzed. If the missing value ratio is less than or equal to a preset ratio, the mean of the corresponding field is used to fill the missing values. If the missing value ratio is greater than the preset ratio, a missing value prediction model is constructed using a random forest algorithm, and the missing value is predicted and filled using the missing value prediction model. Textual missing values ​​are filled uniformly. Based on 3 The principle is to detect outliers in numerical data and replace the detected outliers with the mean of the nearest normal values; the isolated forest algorithm is used to detect outliers in time series data and the detected outliers are corrected by regression analysis or interpolation. The processed multi-source heterogeneous data undergoes data format unification, data type conversion, and normalization.

4. The system according to claim 1, characterized in that, The visualization mapping module specifically includes: The feature analysis unit, connected to the visualization template adjustment unit, is used to extract the data features of the standard data and recommend preliminary visualization templates to users based on the data features and user history behavior through a recommendation algorithm. The visualization template adjustment unit, connected to the feature analysis unit and the data field binding unit, is used to receive user requirements, determine the initial visualization template to be used based on the user requirements, and support users to customize the initial visualization template to be used. The data field binding unit is connected to the visualization template adjustment unit. It is used to receive the adjusted visualization template information and bind the data fields to the attributes of the visualization elements in the visualization template information to complete the conversion of data into visualization graphics.

5. The system according to claim 1, characterized in that, The data fusion module specifically includes: A semantic feature extraction unit, connected to a semantic similarity calculation unit, is used to perform adaptive feature extraction on the visualized data to obtain a unified semantic representation result; A semantic similarity calculation unit, connected to the semantic feature extraction unit and the machine learning fusion unit, is used to calculate the semantic similarity between different data based on the semantic representation results using a cosine similarity algorithm. The machine learning fusion unit, connected to the semantic similarity calculation unit, is used to merge similar data using the DBSCAN density clustering algorithm based on the semantic similarity, and to classify and disambiguate the fused data using the decision tree algorithm.

6. The system according to claim 1, characterized in that, The dynamic ontology construction module specifically includes: An initial ontology construction unit, connected to an incremental learning unit, is used to construct an initial ontology based on the fused data using the OWL language, and to define the classes, attributes, and relationships of the initial ontology; An incremental learning unit, connected to the initial ontology construction unit and the consistency verification unit, is used to monitor data in real time. When the data changes, the Jena framework and ontology mapping algorithm are used to perform semantic matching between the new data and the initial ontology, and the ontology structure is updated or expanded according to the matching results. The consistency verification unit, connected to the incremental learning unit, is used to perform consistency checks on the updated or expanded ontology using the Pellet inference engine, and to automatically roll back or adjust the ontology structure when conflicts are found.

7. The system according to claim 1, characterized in that, The front-end visualization interaction module includes a visualization result rendering unit, an interactive operation processing unit, and a personalization customization unit that are interconnected with each other. The visualization result rendering unit is used to render the data fusion process, ontology structure and data relationship into a front-end visualization interface, and supports adaptive scaling of the front-end visualization interface. The interactive operation processing unit is used to receive user operation information and complete various types of interactions between the user and the system based on the operation information. The personalized customization unit is used to support users in customizing visualization effect parameters and saving the customized visualization effect as a template to the system.

8. A data fusion and knowledge modeling method, characterized in that... include: The data acquisition module collects multi-source heterogeneous data in real time according to the acquisition requirements, and transmits the multi-source heterogeneous data to the data preprocessing module. The multi-source heterogeneous data is received by the data preprocessing module, adaptive preprocessing is performed on the multi-source heterogeneous data to obtain the processed standard data, and the standard data is transmitted to the visualization mapping module. The standard data is received by the visualization mapping module, and the standard data is visualized according to the data characteristics and user needs to obtain visualized data. The visualized data is then sent to the data fusion module. The visualization data is received by the data fusion module, the visualization data is intelligently fused, and the fused data is transmitted to the dynamic ontology construction module. The fused data is received through a dynamic ontology construction module, and an ontology is dynamically constructed based on the fused data using an incremental learning algorithm and a consistency verification mechanism. The front-end visual interaction module provides users with a front-end visual interface and multiple types of interactive interfaces.

9. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the data fusion and knowledge modeling method as described in claim 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores an implementation program for information transmission, which, when executed by a processor, implements the steps of the data fusion and knowledge modeling method as described in claim 8.

Images