Multi-source heterogeneous data fusion processing and intelligent analysis method, device and equipment

Abstract

The invention discloses a multi-source heterogeneous data fusion processing and intelligent analysis method, device and equipment, and relates to the field of data fusion processing.The method comprises the steps that firstly, data in data sources of corresponding types are collected through different collection strategies; performing data preprocessing on the collected data to obtain standard data in a unified format; performing data fusion on the standard data based on the structural features of the data to obtain a fused data set; and determining a target analysis model according to the data analysis target, and performing data analysis on the fused data set. According to the method, various types of data formats can be processed, multi-source heterogeneous data fusion is realized, and through data semantic understanding and a feature-based fusion algorithm, the accuracy and quality of data fusion are improved, and correct data association is established on the semantic level.

Description

Technical Field

[0001] The present invention relates to the field of data fusion processing, and more specifically, to a multi-source heterogeneous data fusion processing and intelligent analysis method, device and equipment. Background Art

[0002] With the rapid development of computer network technology, communication technology, signal detection and other technologies, the degree of informatization in various industries has been continuously improved, the data sources have become increasingly extensive, and the data types have become complex and diverse. People's requirements for the richness, accuracy, and real-time nature of data content can no longer be met by information from a single data source. For example, in the industrial Internet, there are both real-time production data collected by sensors (such as numerical data such as temperature, pressure, and flow), equipment operation logs (text data), and equipment maintenance records (structured or semi-structured data); in the construction of smart cities, there are video data from traffic monitoring cameras, traffic flow sensor data, text information from citizens' feedback, and various statistical report data from urban management departments.

[0003] Multi-source data usually comes from different data sources, different formats, different structures or different feature representation methods, resulting in heterogeneity between data. In addition, the data quality and reliability of different data sources may vary. There may be noise, missing values ​​or errors in multi-source data. Due to uneven data quality and large differences in data formats, it is difficult to directly integrate and process data from different data sources. Traditional data processing methods mainly process a single type of data. When the amount of data is large, it often takes a lot of time and computing resources, and the data processing efficiency is low, which cannot meet the real-time requirements. Summary of the invention

[0004] In view of this, the present invention provides a multi-source heterogeneous data fusion processing and intelligent analysis method, device and equipment, which are used to solve the problem that existing data processing methods only process a single type of data and cannot perform real-time analysis on large amounts of data.

[0005] To achieve the above objectives, the proposed solution is as follows: A multi-source heterogeneous data fusion processing and intelligent analysis method, comprising: Collect data from corresponding types of data sources through different collection strategies; Perform data preprocessing on the collected data to obtain standard data in a unified format; Perform data fusion on the standard data based on the structural characteristics of the data to obtain a fused data set; Determine the target analysis model according to the data analysis objectives, and perform data analysis on the fused data set.

[0006] Preferably, the data source types include: database types, file data, sensor data, real-time message queues and web page data, wherein the database types include relational databases and non-relational databases.

[0007] Preferably, the correspondence between the data source type and the collection strategy includes: If the data source type is a database type, the collection strategy is to collect data using the corresponding query statement; If the data source type is sensor data, the collection strategy is to receive the data sent by the sensor in real time based on the communication protocol; If the data source type is file data, the collection strategy is to collect data through the file reading library corresponding to the file type.

[0008] Preferably, the process of preprocessing the collected data includes: Based on the preset data format conversion framework, the collected data in different formats are converted into data in a unified format; Clean and repair the data in a unified format; Perform semantic annotation on the repaired data to obtain standard data.

[0009] Preferably, it also includes: The data analysis results are displayed.

[0010] Preferably, the process of fusing the standard data based on the structural features includes: If the data is homogeneous, the data fusion algorithm is based on the primary key or unique identifier; If the data is heterogeneous, the data fusion algorithm is to fuse the data based on the similarity between the semantics and features.

[0011] A multi-source heterogeneous data fusion processing and intelligent analysis device, further comprising: A data collection unit collects data from corresponding types of data sources through different collection strategies; A data preprocessing unit performs data preprocessing on the collected data to obtain standard data in a unified format; A data fusion unit performs data fusion on standard data based on the structural characteristics of the data to obtain a fused data set; The intelligent analysis unit determines the target analysis model according to the data analysis target and performs data analysis on the fused data set.

[0012] A multi-source heterogeneous data fusion processing and intelligent analysis device, comprising: a memory and a processor; The memory is used to store programs; The processor is used to execute the program to implement the various steps of the aforementioned multi-source heterogeneous data fusion processing and intelligent analysis method.

[0013] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects: (1) The multi-source heterogeneous data fusion processing and intelligent analysis method provided by the present invention first collects data from corresponding types of data sources through different collection strategies; performs data preprocessing on the collected data to obtain standard data in a unified format; performs data fusion on the standard data based on the structural characteristics of the data to obtain a fused data set; determines the target analysis model according to the data analysis target, and performs data analysis on the fused data set. The present invention can process various types of data formats and realize multi-source heterogeneous data fusion. Through data semantic understanding and feature-based fusion algorithms, it not only improves the accuracy and quality of data fusion, but also establishes correct data associations at the semantic level.

[0014] (2) The multi-source heterogeneous data fusion processing and intelligent analysis method provided by the present invention adopts a parallel processing architecture, fully utilizing multi-core CPUs and distributed computing resources in various links such as data acquisition, preprocessing, fusion and analysis, greatly improving the data processing speed. It can quickly process massive multi-source heterogeneous data, improve data processing efficiency, and better meet real-time requirements.

[0015] (3) The multi-source heterogeneous data fusion processing and intelligent analysis method provided by the present invention automatically matches the appropriate model according to the data characteristics and analysis objectives, avoiding the blindness of manual model selection and improving the accuracy and reliability of the analysis results. At the same time, through the model evaluation and optimization mechanism, the performance of the analysis model is continuously improved, further improving the efficiency of data analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0017] Figure 1 A flow chart of a multi-source heterogeneous data fusion processing and intelligent analysis method provided by an embodiment of the present invention; Figure 2 A schematic diagram of the structure of a multi-source heterogeneous data fusion processing and intelligent analysis device provided by an embodiment of the present invention; Figure 3 A hardware structure block diagram of a multi-source heterogeneous data fusion processing and intelligent analysis device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] First, combine Figure 1 A multi-source heterogeneous data fusion processing and intelligent analysis method provided by an embodiment of the present invention is introduced. Figure 1 As shown, the method includes: Step S01: collect data from corresponding types of data sources through different collection strategies.

[0020] Specifically, by developing a variety of data source adapters, they are used to connect different types of data sources, including database types, file data (such as text files, log files, spreadsheet files, etc.), sensor data (such as data accessed by sensor network protocols such as ZigBee and MQTT), real-time message queues (such as Kafka, ActiveMQ, etc.) and web page data (such as web page data obtained through web crawlers, etc.), among which database types include relational databases (such as MySQL, Oracle, etc.) and non-relational databases (MongoDB, Cassandra, etc.). According to the characteristics of different types of data sources, corresponding collection strategies are adopted. For example: for database data sources, incremental data is regularly obtained through SQL query statements, etc.; for sensor data, sensor data push is monitored in real time through sensor network protocols and cached; for web page data, updated content is regularly captured according to the set crawling rules; for different types of files, corresponding file reading libraries are used for data collection.

[0021] Step S02, preprocessing the collected data to obtain standard data in a unified format.

[0022] Specifically, the data is processed through format conversion, cleaning, repair and semantic understanding, and finally standard data in a unified format is obtained.

[0023] Data format conversion: Design a general data format conversion framework to convert the collected data in various formats into a unified intermediate format. Based on the preset data format conversion framework, convert the collected data in different formats into data in a unified format. The unified format can be a JSON format or a custom structured format. When processing structured data such as user information data from different databases, it can be uniformly converted into an object format containing attributes such as user ID, name, age, gender, registration time, etc. For unstructured data such as user comment text data, the TF-IDF algorithm can be used to extract feature vectors and record the corresponding related metadata (such as comment time, product, etc.), and the feature vectors and related metadata are encapsulated and stored.

[0024] For structured data, it can be converted into an object in an intermediate format through mapping rules. For example, structured data can be database query results or Excel table data, etc., and the fields in the database table are mapped to object attributes in a unified format, and the date field is converted into a unified date format (such as ISO 8601 format). For unstructured data (such as text, images, audio, etc.), feature extraction technology is used to convert it into a feature vector representation, and it is encapsulated into an intermediate format together with related metadata. For example: For text data, natural language processing technology can be used for feature extraction, and the bag of words model or TF-IDF (term frequency-inverse document frequency) algorithm can be used to convert the text into a vector representation, and metadata such as keywords and entities in the text are extracted at the same time, and encapsulated into an object in a unified format together with the vector. Among them, feature extraction of text data can be performed through the bag of words model, Chinese language model (N-gram) or word embedding; for image data, image recognition technology is used to extract image features (such as color histogram, texture features, etc. of the image), and it is converted into a vector form and stored together with image-related metadata (such as image name, shooting time, etc.). Among them, feature extraction of image data can be performed through convolutional neural networks, autoencoders, Haar features, etc.

[0025] Data cleaning and repair: First, data cleaning algorithms can be used to clean and repair data. Through data cleaning algorithms, missing values, erroneous values, and duplicate values ​​in the data are identified and processed. At the same time, data consistency checks are performed to ensure the logical consistency of related data in different data sources. Among them, duplicate values ​​can be identified by the hash value or key attribute value of the data record; erroneous values ​​can be detected and corrected according to the error value detection rules set according to the data value range, data type, and business rules. For example: for age data, a reasonable value range is set (such as 0-120 years old), and values ​​outside this range are considered erroneous values.

[0026] Then, the outliers are repaired. When the outliers are missing values, if the missing values ​​are numerical data, they can be repaired by using mean filling, median filling or model-based filling methods. Among them, if the data distribution is relatively uniform, mean filling can be used; if the data is skewed, median filling can be used. If the missing value is text data, it can be supplemented according to context or domain knowledge. For records with serious missing data, they can be deleted according to certain rules, that is, data with a missing value ratio exceeding a certain threshold is deleted. When the outliers are wrong values, they can be corrected by neighboring value replacement, statistical model-based correction or manual intervention. When the outliers are duplicate values, duplicate records can be removed by data hashing or comparison algorithms. Keep one of the records and delete the other duplicate records. During the processing, pay attention to the timeliness of the data and keep the latest records. For example, in an e-commerce sales data set, it is detected that the prices of some products are negative (error values). At this time, the median can be used to correct them based on the historical price of the product and the price distribution of similar products. At the same time, some duplicate order records are found. The duplicate records can be removed by comparing attributes such as order number and order time. For a small amount of missing customer address information, it can be supplemented based on the customer's historical order address.

[0027] Data semantic annotation and understanding The repaired data is semantically annotated based on the pre-built semantic knowledge base to obtain standard data. The data is semantically annotated and understood through natural language processing technology (such as lexical analysis, syntactic analysis, semantic parsing, etc.) and machine learning algorithms (such as classification algorithms, clustering algorithms, etc.). For example, keywords in text data are extracted and classified to identify the semantic category to which they belong; for fields in structured data, semantic interpretation is performed based on the semantic knowledge base to determine their meaning and relationship.

[0028] Collect and organize semantic information such as terms, concepts, data format definitions, data relationships, etc. related to the application field to build a semantic knowledge base. For example, in the medical field, build a knowledge base containing terms such as disease names, symptoms, examination items, treatment methods, and the relationships between these terms (such as the relationship between diseases and symptoms, the diagnostic relationship between examination items and diseases, etc.).

[0029] The perfect data cleaning and repair function effectively solves the problems of missing, error and duplication in the data, and improves the accuracy and completeness of the data. Data semantic annotation and understanding help to find logical inconsistencies in the data, further ensuring the data quality. Among them, high-quality data provides strong support for accurate analysis and decision-making.

[0030] Step S03: performing data fusion on the standard data based on the structural characteristics of the data.

[0031] Specifically, based on the semantic information and features of the data, a suitable data fusion algorithm is used. For homogeneous data (such as data from different databases but with the same structure), a simple data merging algorithm is used for fusion, such as fusion based on primary keys or unique identifiers; for heterogeneous data, fusion is performed based on the semantic association of the data and the similarity between the features. For example, by calculating the similarity between the data, such as cosine similarity and Euclidean distance, similar data is associated and fused to generate a fused data set.

[0032] During the fusion process, the credibility and weight of the data are considered at the same time. Weights are assigned to data based on factors such as the reliability of the data source and the freshness of the data. The more reliable and fresher the data source, the greater the weight assigned. For example, the total weight of reliability and freshness is set to 0.5, and the freshness weight and reliability weight are assigned according to the freshness and reliability of the data source. For example, if the freshness weight of a data source is 0.2 and the reliability weight is 0.1, the weight assigned to the data is 0.3. Data with high credibility has a greater influence on the fusion results, thereby improving the quality of the fused data.

[0033] Step S04, determining a target analysis model according to a data analysis target, and performing data analysis on the fused data set.

[0034] Specifically, according to the characteristics of the data and the data analysis objectives, select a suitable analysis model from the pre-built data analysis model library or combine multiple models for analysis. For example, for predictive analysis tasks, select a suitable machine learning regression model; for data classification tasks, use a classification model for processing; for complex data relationship mining tasks, combine models such as association rule mining and cluster analysis. Evaluate and optimize the analysis results. During model training, the performance of the analysis model is evaluated by cross-validation, accuracy evaluation and other methods, and the model parameters are adjusted or a more suitable model is selected based on the evaluation results to improve the accuracy and reliability of the analysis results. During the training process, cross-validation or accuracy evaluation methods are used to evaluate the performance of each model in different data analysis tasks. Adjust the model parameters according to the accuracy in the evaluation results to improve the accuracy of the model analysis. The data analysis objectives and data characteristics suitable for each model can also be determined based on the evaluation results. When applying, a model with high accuracy can be selected for data analysis based on the characteristics of the data and the data analysis objectives.

[0035] The data analysis model library contains various data analysis algorithms and models, such as statistical analysis models (mean, variance, correlation analysis, etc.), machine learning models (classification models such as decision trees, neural networks, clustering models such as K-Means, regression models such as linear regression, etc.), data mining models (association rule mining, sequence pattern mining, etc.) and deep learning models (such as convolutional neural networks for image analysis, recurrent neural networks for time series analysis, etc.).

[0036] Next, the embodiment of the present invention introduces the corresponding relationship between different data source types and collection strategies through examples, and the corresponding relationship is as follows: If the data source type is a database type, the collection strategy is to collect data using the corresponding query statement. When the relational database type is MySQL, use the JDBC (Java Database Connectivity) driver to establish a connection, configure the database connection parameters (such as URL, user name, password, etc.), and finally collect data using the written SQL query statement. For example, in an e-commerce data analysis project, you can connect to the MySQL database through JDBC at regular intervals every day to query the new order data, product inventory data, etc. of the day, as well as the order data and customer information in the sales database. You can set query conditions according to business needs (such as querying orders in a specific time period, customer data in a specific region, etc.). When the non-relational database is MongoDB, use the corresponding MongoDB driver to connect to the database and collect data using MongoDB's query syntax.

[0037] If the data source type is sensor data, the collection strategy is to receive the data sent by the sensor in real time based on the communication protocol. For sensor networks using the ZigBee protocol, use the ZigBee communication library to establish a connection with the sensor node and receive the data sent by the sensor in real time (such as temperature, humidity, light intensity, etc.). For sensors that support the MQTT protocol, use the MQTT client library to subscribe to the topics published by the sensor to collect data and cache it. For example, in a smart home system, connect to the indoor temperature and humidity sensors through the MQTT client to collect environmental data in real time.

[0038] If the data source type is file data, the collection strategy is to collect data through the file reading library corresponding to the file type. For text files, you can use Java's BufferedReader or Python's open function to read line by line; for CSV files, you can use Python's csv module for parsing; for Excel files, you can use Java's Apache POI library or Python's openpyxl library for data collection. After the file data is read, useful information can be extracted and converted into a unified format based on the structure and content of the file. For example: In an enterprise financial data analysis project, the financial statement data in Excel format provided by the finance department is read every month, and the income, expenditure, profit and other data are extracted and collected, and converted into a format that the system can process.

[0039] If the data source type is a real-time message queue, data can be collected through the corresponding data collection interface. For Kafka message queues, you can use the Kafka Consumer API to subscribe to related topics and receive real-time message data. Ensure reliable data reception by configuring the parameters of the Kafka consumer (such as consumer groups, offsets, etc.). Message data may come from different data sources, such as real-time data push and log data forwarding from other systems. For example: In a big data platform, real-time log data from multiple business systems can be received through Kafka. These log data contain information such as system operation status and user operation records.

[0040] The multi-source heterogeneous data fusion processing and intelligent analysis method provided by the embodiment of the present invention sets the corresponding collection strategy according to the characteristics of the data source type. The processing method provided by the embodiment of the present invention can process various types of data sources and data formats, and realize the real multi-source heterogeneous data fusion. Whether it is structured database data, semi-structured log files or unstructured images and text data, it can be effectively fused to provide a comprehensive data foundation for subsequent comprehensive analysis.

[0041] In order to allow users to better understand the analysis results and formulate corresponding strategies based on the analysis results, the embodiment of the present invention, based on the multi-source heterogeneous data fusion processing and intelligent analysis method of the above embodiment, also includes: The data analysis results are displayed.

[0042] Specifically, according to user needs, the analysis results are displayed in a variety of forms to facilitate users to intuitively understand and use the analysis results, such as: visual charts such as bar charts, line charts, pie charts or scatter charts, reports in Excel or PDF formats, interactive interfaces such as web pages or mobile applications, and data interfaces such as RESTful APIs.

[0043] The analysis results can be applied to actual business scenarios. For example, market trend analysis, risk assessment and other analysis results can provide support for corporate decision-making, and fault prediction and maintenance scheduling can be performed based on equipment operation data analysis to optimize industrial production processes.

[0044] The flexible architecture of the present invention enables the present invention to be easily applied to different fields and industries, such as industrial manufacturing, financial services, medical care, smart cities, etc. According to the needs of different industries, the specific processes and algorithms of data collection, preprocessing, analysis and display can be customized to meet diverse business needs, with strong versatility and scalability.

[0045] A multi-source heterogeneous data fusion processing and intelligent analysis device provided in an embodiment of the present invention is described below. The multi-source heterogeneous data fusion processing and intelligent analysis device described below and the multi-source heterogeneous data fusion processing and intelligent analysis method described above can refer to each other.

[0046] First combine Figure 2 , introduces the multi-source heterogeneous data fusion processing and intelligent analysis device, such as Figure 2 As shown, the multi-source heterogeneous data fusion processing and intelligent analysis device may include: The data collection unit 100 collects data from corresponding types of data sources through different collection strategies; it includes multiple data source connectors, which are responsible for establishing connections with various data sources and collecting data, and transmitting the collected data to the data preprocessing unit 200.

[0047] The data preprocessing unit 200 performs data preprocessing on the collected data to obtain standard data in a unified format; after the data preprocessing unit 200 performs format conversion, cleaning, repairing and semantic understanding on the data, the processed data is passed to the data fusion unit 300.

[0048] The data fusion unit 300 performs data fusion on the standard data based on the structural characteristics of the data to obtain a fused data set; the data fusion unit 300 sends the fused data set to the intelligent analysis unit 400.

[0049] The intelligent analysis unit 400 determines the target analysis model according to the data analysis target and performs data analysis on the fused data set. The intelligent analysis unit 400 uses the model in the data analysis model library for analysis.

[0050] The multi-source heterogeneous data fusion processing and intelligent analysis device provided in the embodiment of the present invention can be applied to multi-source heterogeneous data fusion processing and intelligent analysis equipment. Figure 3 The hardware structure diagram of the multi-source heterogeneous data fusion processing and intelligent analysis equipment is shown in FIG. Figure 3 , the hardware structure of the device may include: at least one processor 1, at least one communication interface 2, at least one memory 3 and at least one communication bus 4; In the embodiment of the present invention, the number of the processor 1, the communication interface 2, the memory 3, and the communication bus 4 is at least one, and the processor 1, the communication interface 2, and the memory 3 communicate with each other through the communication bus 4; The processor 1 may be a central processing unit CPU, or an application-specific integrated circuit ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present invention, etc.; The memory 3 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), etc., such as at least one disk memory; The memory stores a program, and the processor can call the program stored in the memory, wherein the program is used to implement each processing flow in the aforementioned multi-source heterogeneous data fusion processing and intelligent analysis solution; An embodiment of the present invention further provides a storage medium, which can store a program suitable for execution by a processor, wherein the program is used to implement each processing flow in the aforementioned multi-source heterogeneous data fusion processing and intelligent analysis solution.

[0051] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0052] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0053] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multi-source heterogeneous data fusion processing and intelligent analysis method, characterized in that: include: Collect data from corresponding types of data sources through different collection strategies; Perform data preprocessing on the collected data to obtain standard data in a unified format; Perform data fusion on the standard data based on the structural characteristics of the data to obtain a fused data set; Determine the target analysis model according to the data analysis objectives, and perform data analysis on the fused data set.

2. The multi-source heterogeneous data fusion processing and intelligent analysis method according to claim 1 is characterized in that: Data source types include: database types, file data, sensor data, real-time message queues, and web page data, where database types include relational databases and non-relational databases.

3. The multi-source heterogeneous data fusion processing and intelligent analysis method according to claim 2 is characterized in that: The correspondence between data source types and collection strategies, including If the data source type is a database type, the collection strategy is to collect data using the corresponding query statement; If the data source type is sensor data, the collection strategy is to receive the data sent by the sensor in real time based on the communication protocol; If the data source type is file data, the collection strategy is to collect data through the file reading library corresponding to the file type.

4. The multi-source heterogeneous data fusion processing and intelligent analysis method according to claim 1 is characterized in that: The process of preprocessing the collected data includes: Based on the preset data format conversion framework, the collected data in different formats are converted into data in a unified format; Clean and repair the data in a unified format; Perform semantic annotation on the repaired data to obtain standard data.

5. The multi-source heterogeneous data fusion processing and intelligent analysis method according to claim 1 is characterized in that: Also includes: The data analysis results are displayed.

6. The multi-source heterogeneous data fusion processing and intelligent analysis method according to claim 1 is characterized in that: The process of fusing standard data based on structural features includes: If the data is homogeneous, the data fusion algorithm is based on the primary key or unique identifier; If the data is heterogeneous, the data fusion algorithm is to fuse the data based on the similarity between the semantics and features.

7. A multi-source heterogeneous data fusion processing and intelligent analysis device, characterized in that: Also includes: A data collection unit collects data from corresponding types of data sources through different collection strategies; A data preprocessing unit performs data preprocessing on the collected data to obtain standard data in a unified format; A data fusion unit performs data fusion on standard data based on the structural characteristics of the data to obtain a fused data set; The intelligent analysis unit determines the target analysis model according to the data analysis target and performs data analysis on the fused data set.

8. A multi-source heterogeneous data fusion processing and intelligent analysis device, characterized in that: include: Memory and processor; The memory is used to store programs; The processor is used to execute the program to implement each step of the multi-source heterogeneous data fusion processing and intelligent analysis method as described in any one of claims 1 to 6.

Images