A Big Data Fusion Computing Method and System

Abstract

The present invention discloses a big data fusion calculation method and system. The method includes: receiving input data from a multi-source heterogeneous data universe, where each data universe provides data with different dimensions and structures; converting the input data into a unified data lattice, and extracting cross-dimensional feature data from the data lattice; performing data weaving processing on the cross-dimensional feature data, and outputting the data after the weaving processing in a hyper-dimensional form. By using the embodiments of the present invention, it is possible to have the flexibility to process different data types, and at the same time, extract effective information from complex data relationships, and generate a new data form, providing more in-depth data insights for various fields.

Description

Technical Field

[0001] The present invention belongs to the technical field of big data, and particularly relates to a big data fusion calculation method and system. Background Art

[0002] With the continuous development of information technology, the amount of data has shown an explosive growth, and the sources have become more and more diverse. The massive data generated in all walks of life contains extremely rich information and value. Effectively fusing and processing this big data can not only provide more comprehensive decision-making support, but also bring new business insights and scientific discoveries.

[0003] Traditional data processing methods often focus on the analysis and mining of data from a single source, and it is difficult to handle multi-source heterogeneous data from different data universes. These data universes may involve different fields, different formats and structures, as well as different data quality and integrity. In this context, traditional data analysis methods are not sufficient to process and fuse information from these heterogeneous data sources. Therefore, the big data fusion calculation method has become a key technology to solve this problem. Summary of the Invention

[0004] The object of the present invention is to provide a big data fusion calculation method and system to solve the deficiencies in the prior art, which can have the flexibility to process different data types, and at the same time can extract effective information from complex data relationships and generate a new data form to provide more in-depth data insights for various fields.

[0005] An embodiment of the present application provides a big data fusion calculation method, and the method includes:

[0006] Receiving input data from multi-source heterogeneous data universes, where each data universe provides data with different dimensions and structures;

[0007] Converting the input data into a unified data lattice, and extracting cross-dimensional feature data from the data lattice;

[0008] Performing data weaving processing on the cross-dimensional feature data, and outputting the data after the weaving processing in a hyper-dimensional form.

[0009] Optionally, the converting the input data into a unified data lattice includes:

[0010] Regarding the input data as nodes in a graph, and using graph embedding technology to map the nodes into a multi-dimensional data lattice structure to obtain a multi-dimensional data lattice based on the graph structure, where the edges of the graph represent the relationships between the nodes.

[0011] Optionally, the extracting cross-dimensional feature data from the data lattice includes:

[0012] Perform persistent homology analysis on the multi-dimensional data lattice, and represent the analysis result of persistent homology as a topological summary of the data;

[0013] Construct a topological feature vector according to the topological summary, and combine the topological feature vectors of different scales to generate a multi-dimensional feature representation.

[0014] Optionally, the data weaving process is performed on the cross-dimensional feature data, and the data after the weaving process is output in a hyper-dimensional form, including:

[0015] Map the cross-dimensional feature data into a pre-constructed symbol system, where each feature point or feature set in the cross-dimensional feature data is represented by one or a group of symbols;

[0016] Perform spatial mapping on the cross-dimensional feature data to generate a hyper-dimensional representation space;

[0017] Weave the symbols in the symbol system in the hyper-dimensional representation space to generate and output a hyper-dimensional form.

[0018] Another embodiment of the present application provides a big data fusion computing system, and the system includes:

[0019] A receiving module, configured to receive input data from a multi-source heterogeneous data universe, where each data universe provides data of different dimensions and structures;

[0020] A conversion module, configured to convert the input data into a unified data lattice, and extract cross-dimensional feature data from the data lattice;

[0021] A weaving module, configured to perform a data weaving process on the cross-dimensional feature data, and output the data after the weaving process in a hyper-dimensional form.

[0022] Another embodiment of the present application provides a storage medium, in which a computer program is stored, and the computer program is configured to execute the method described in any one of the above when running.

[0023] Another embodiment of the present application provides an electronic device, including a memory and a processor, a computer program is stored in the memory, and the processor is configured to run the computer program to execute the method described in any one of the above.

[0024] Compared with the prior art, a big data fusion calculation method provided by the present invention receives input data from a multi-source heterogeneous data universe, where each data universe provides data with different dimensions and structures; converts the input data into a unified data lattice, extracts cross-dimensional feature data from the data lattice; performs data weaving processing on the cross-dimensional feature data, and outputs the processed data in a hyper-dimensional form, so as to be flexible in processing different data types, extract effective information from complex data relationships at the same time, and generate a new data form, providing deeper data insights for various fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a hardware structure block diagram of a computer terminal for a big data fusion calculation method provided by an embodiment of the present invention;

[0026] Figure 2 is a schematic flowchart of a big data fusion calculation method provided by an embodiment of the present invention;

[0027] Figure 3 is a schematic structural diagram of a big data fusion calculation system provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0028] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0029] An embodiment of the present invention first provides a big data fusion calculation method, which can be applied to an electronic device, such as a computer terminal, specifically, an ordinary computer, etc.

[0030] The following takes running on a computer terminal as an example to describe it in detail. Figure 1 is a hardware structure block diagram of a computer terminal for a distributed energy storage management method provided by an embodiment of the present invention. As Figure 1 shown, the computer terminal may include one or more ( Figure 1 only one is shown in the figure) processors 102 (the processor 102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data. Optionally, the above computer terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above computer terminal. For example, the computer terminal may further include more or fewer components than Figure 1 shown in the figure, or have a different configuration from Figure 1 shown in the figure.

[0031] The memory 104 can be used to store software programs and modules of application software, such as program instructions / modules corresponding to the distributed energy storage management method in the embodiments of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implements the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely disposed relative to the processor 102, and these remote memories can be connected to the computer terminal through a network. Examples of the above network include but are not limited to the Internet, intranet, local area network, mobile communication network, and combinations thereof.

[0032] The transmission device 106 is used to receive or send data via a network. Specific examples of the above network may include a wireless network provided by a communication provider of a computer terminal. In one instance, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station and thus communicate with the Internet. In one instance, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0033] See Figure 2 , embodiments of the present invention provide a big data fusion calculation method, which may include the following steps:

[0034] S201, receive input data from a multi-source heterogeneous data universe, where each data universe provides data of different dimensions and structures;

[0035] Specifically, the following is a specific implementation manner to implement this step:

[0036] Step 1: Data source identification and access

[0037] 1. Define the data universe directory: Establish a global directory for identifying and classifying different data sources. This directory will contain meta-information of all data universes, such as data types, data structures, update frequencies, and data quality indicators.

[0038] 2. Source heterogeneity analysis: Conduct in-depth analysis on each data universe to determine its data format (such as CSV, JSON, XML), data access protocol (such as HTTP REST API, FTP), the degree of data structuring (structured, semi-structured, unstructured), and specific data patterns.

[0039] 3. Data Access Pipeline Establishment: Based on the analysis results, establish a data access pipeline for each data universe. The role of the pipeline is to automatically introduce data from the original source into the fusion system, and at the same time, the pipeline needs to flexibly support different communication protocols and data formats.

[0040] Step 2: Data Preprocessing and Normalization

[0041] 4. Data Cleaning: Clean the introduced data, including removing redundancy, correcting errors, filling in missing values, etc. In this step, specific cleaning rules need to be designed for data with different dimensions and structures.

[0042] 5. Normalization and Transformation: Convert heterogeneous data into a unified format and structure so that it can be represented by a unified data lattice. This may involve data type conversion, timestamp unification, encoding adjustment, etc.

[0043] Step 3: Data Semantic Integration

[0044] 6. Establish a Semantic Model: Establish a unified semantic hierarchical model for the data in the data universe. This model will define the meaning of data from different data sources and the relationships between them, enabling data from different dimensions to be aligned at the semantic level.

[0045] 7. Entity Resolution and Linking: Use entity recognition and linking techniques to match key entities (such as people, organizations, locations) in the data with semantic concepts in the model to ensure semantic consistency of data from different sources.

[0046] Step 4: Metadata Management and Data Lattice Construction

[0047] 8. Establish a Metadata Warehouse: Create a metadata management system to store descriptive information about the data, including data sources, dimensions, structures, semantic definitions, and data quality information.

[0048] 9. Dynamic Data Lattice Generation: Dynamically generate a data lattice based on the existing metadata and semantic model. The data lattice should not only include the content of the data but also be able to reflect the relationships between the data so that it can support subsequent cross-dimensional feature extraction.

[0049] Through the above steps, the input data received from the multi-source heterogeneous data universe is gradually converted into a unified format with deep integration and semantic consistency, which will subsequently provide a basis for other parts of the fusion method (feature extraction, data weaving, and hyperdimensional morphological output). This process takes into account various features of the data and improves the accuracy and efficiency of data fusion through intelligent processing and metadata management.

[0050] S202. Convert the input data into a unified data lattice and extract cross-dimensional feature data from the data lattice.

[0051] Specifically, the input data can be regarded as nodes in a graph, and the graph embedding technique is used to map the nodes into a multi-dimensional data lattice structure to obtain a graph-based multi-dimensional data lattice, where the edges of the graph represent the relationships between the nodes. A specific implementation method can be as follows:

[0052] 1. Data preprocessing: Use an autoencoder for deep feature learning (preferably)

[0053] Autoencoder standardization: Use an autoencoder in deep learning to perform deep feature learning on the received multi-source heterogeneous data. Through the autoencoder, not only can data standardization be achieved, but also deep-level features of the data can be learned, which helps for more effective subsequent data mapping and feature extraction.

[0054] Noise and outlier processing: Identify and remove noise data and outliers through the reconstruction error of the autoencoder. This method can understand the data at the deep learning level, thus more effectively identifying and removing outliers that do not conform to the overall characteristics of the data set.

[0055] 2. Data mapping: Construction of a data lattice based on graph embedding

[0056] Graph embedding technique: Regard the standardized data as nodes in a graph, and map these data nodes into a multi-dimensional data lattice structure through the graph embedding technique. The graph embedding technique can effectively preserve the relative positions and relationships between data points, so that the essential structure and connections of the data can be reflected even in a high-dimensional space.

[0057] Lattice structure design: Design a graph-based multi-dimensional data lattice, and use the edges of the graph to represent the relationships between data points, with nodes representing data points. Such a structure not only helps to preserve the original structural information of the data, but also can be flexibly extended to high-dimensional data.

[0058] Specifically, persistent homology analysis can be performed on the multi-dimensional data lattice, and the analysis results of persistent homology are represented as the topological summary of the data; construct a topological feature vector according to the topological summary, and combine topological feature vectors at different scales to generate a multi-dimensional feature representation. A specific implementation method can be as follows:

[0059] Step 1: Topological representation of the data lattice

[0060] Persistent homology computation: Perform persistent homology analysis on the data lattice, which is a major tool for studying the shape of data. By establishing a multi-scale filtration sequence and observing the emergence and disappearance of data features (such as connected components, loop structures, holes, etc.) at different scales, the topological features of the data are captured.

[0061] Establish topological summaries: Represent the results of persistent homology as topological summaries, which can be regarded as topological summaries of the data, capturing the topological changes of the data as the scale varies.

[0062] Step 2: Cross-dimensional feature extraction

[0063] Feature vector construction: Construct feature vectors based on the topological summaries, which can be achieved by transforming the topological summaries into a vector form suitable for machine learning. For example, using vectorization techniques such as Persistence Landscape or Kernel methods.

[0064] Multi-scale feature fusion: Combine topological feature vectors from different scales to form a comprehensive multi-dimensional feature representation, reflecting the topological structure information of each dimension in the data lattice.

[0065] Step 3: Feature dimension reduction (preferred)

[0066] Feature selection: Since this method usually generates a large number of topological features, feature selection is required to eliminate redundant and unimportant features and reduce the computational complexity.

[0067] Application of dimension reduction techniques: Apply dimension reduction techniques (such as t-SNE or UMAP) to process the feature vectors in order to more effectively identify and visualize patterns and groups in the data.

[0068] In another implementation in practical applications, it may specifically include the following:

[0069] Step 1: Data acoustic encoding

[0070] Signal conversion: First, each data point in the data lattice can be converted into an acoustic signal. This conversion can be achieved by assigning different frequencies to different features, mapping the data point values to the properties of sound waves, such as frequency, amplitude, or phase.

[0071] Synthesize data sounds: Combine the converted sound waves according to the positions of the data points in the data lattice. Each dimension can represent a sound channel (similar to the concept of left and right channels in stereo). Finally, a "sound" representation of the multi-dimensional data set can be obtained.

[0072] Step 2: Acoustics-based feature extraction

[0073] Spectrum Analysis: Conduct spectrum analysis on the synthesized data sound to identify unique frequency patterns present in the data. Spectrum analysis can reveal potential connections between different data dimensions. For example, resonance and harmonic relationships of frequencies may indicate the internal structure of the data.

[0074] Acoustic Pattern Recognition: Use acoustic pattern recognition techniques, such as voiceprint recognition algorithms, to extract unique features of the data sound. These algorithms are typically used to identify complex features in human voices but can also be applied to identify features of any sound signal.

[0075] Step 3: Statistical Analysis and Classification of Features

[0076] Statistical Analysis: Conduct statistical analysis on the features extracted from the acoustic analysis to determine which features are significant. This may involve calculating statistics such as the mean, variance, skewness, and kurtosis of the features.

[0077] Application of Classification Algorithms: Utilize classification algorithms in machine learning (such as support vector machines or random forests) to analyze the distribution of features and classify the data. This step can help understand the differences in acoustic features among different classes of data.

[0078] Step 4: Feature Dimension Optimization and Selection

[0079] Dimension Optimization: A large number of acoustic features may be generated, so use dimension optimization methods to reduce the number of features. Methods such as principal component analysis (PCA) can remove redundancy and retain the most important features.

[0080] Feature Selection: Through feature selection algorithms, such as minimum redundancy maximum relevance feature selection (mRMR), select the features most relevant to data classification.

[0081] Step 5: Integration and Output of Cross-Dimensional Feature Data

[0082] Feature Data Integration: Combine the acoustic features with other features in the original data to form a comprehensive dataset that includes the original data and cross-dimensional acoustic features.

[0083] Output: Output the integrated feature dataset for further analysis or as input for other data processing and machine learning tasks.

[0084] This method combines the characteristics of traditional data processing and acoustic analysis, providing a novel perspective for the extraction of cross-dimensional feature data. The physical properties of acoustic signals allow data to be understood and processed in a form different from pure mathematical or statistical methods. Pattern recognition at the perceptual level brings new possibilities for data lattice analysis.

[0085] S203. Perform data weaving processing on the cross-dimensional feature data and output the data after the weaving processing in a hyperdimensional form.

[0086] Specifically, the cross-dimensional feature data can be mapped into a pre-constructed symbol system, where each feature point or feature set in the cross-dimensional feature data is represented by one or a group of symbols; perform a spatial mapping on the cross-dimensional feature data to generate a hyperdimensional representation space; weave the symbols in the symbol system in the hyperdimensional representation space to generate and output in a hyperdimensional form. One implementation method can be as follows:

[0087] Step 1: Establish a hyperdimensional representation space

[0088] Define the symbol system: First, define a symbol system, and each symbol corresponds to a feature point or feature set in the data lattice. These symbols will serve as the basic building blocks in the subsequent weaving process.

[0089] Multidimensional scaling mapping: Use the multidimensional scaling algorithm (MDS) to perform a spatial mapping on the obtained cross-dimensional feature data to generate a hyperdimensional representation space. This step utilizes the similarity or distance between features to maintain the topological structure between points.

[0090] Step 2: Symbol encoding of feature data

[0091] Mapping from feature to symbol: Map the cross-dimensional feature data into the symbol system, and each feature point or set is represented by one or a group of symbols.

[0092] Neuro-symbolic learning: Use a neural network to learn and optimize the association rules between symbols, and these rules will be used to guide the weaving process. Here, the symbols not only represent data points but may also contain information about how features interact with each other.

[0093] Step 3: Perform data weaving processing

[0094] Formulation of symbol execution rules: Based on the association rules learned in the previous step, formulate a set of symbol execution rules. These rules determine how the data is intertwined in the hyperdimensional space.

[0095] Implementation of the weaving algorithm: Execute the weaving algorithm to construct complex hyperdimensional forms by gradually superimposing and combining symbols. This may involve iteratively merging symbols into new symbols at a higher level, representing more complex feature relationships.

[0096] Step 4: Visualization and optimization of the hyperdimensional form

[0097] Hyperdimensional Morphology Visualization: Using visualization tools to convert hyperdimensional morphology into visual images, such as multi-dimensional geometric figures or symbolic network diagrams, enabling humans to intuitively understand hyperdimensional data structures.

[0098] Iterative Optimization: Based on the visualization results, iteratively optimize the symbol system, neural network model, and weaving rules to refine the hyperdimensional morphology structure.

[0099] Step 5: Output of Hyperdimensional Morphology Data

[0100] Data Formatting: Determine an output format suitable for representing hyperdimensional morphology data, which can be a graphic file, data file, or specially designed data structure.

[0101] Output Encapsulation: Encapsulate the optimized hyperdimensional morphology, embed it into the selected output format, and output and store it for subsequent use and analysis.

[0102] It can be seen that by receiving input data from a multi-source heterogeneous data universe, where each data universe provides data of different dimensions and structures; converting the input data into a unified data lattice, extracting cross-dimensional feature data from the data lattice; performing data weaving processing on the cross-dimensional feature data, and outputting the woven data in hyperdimensional morphology, it is possible to have the flexibility to handle different data types, simultaneously extract valid information from complex data relationships, and generate a new data form, providing more in-depth data insights for various fields.

[0103] Another embodiment of the present invention provides a big data fusion computing system. Refer to Figure 3 , the system may include:

[0104] A receiving module 301, configured to receive input data from a multi-source heterogeneous data universe, where each data universe provides data of different dimensions and structures;

[0105] A conversion module 302, configured to convert the input data into a unified data lattice and extract cross-dimensional feature data from the data lattice;

[0106] A weaving module 303, configured to perform data weaving processing on the cross-dimensional feature data and output the woven data in hyperdimensional morphology.

[0107] It can be seen that by receiving input data from a multi-source heterogeneous data universe, where each data universe provides data with different dimensions and structures; converting the input data into a unified data lattice, extracting cross-dimensional feature data from the data lattice; performing data weaving processing on the cross-dimensional feature data, and outputting the data after the weaving processing in a hyper-dimensional form, it is possible to have the flexibility to process different data types, while being able to extract effective information from complex data relationships, and generate a new data form, providing more in-depth data insights for various fields.

[0108] An embodiment of the present invention also provides a storage medium, in which a computer program is stored, and the computer program is set to execute the steps in any one of the above method embodiments when running.

[0109] Specifically, in this embodiment, the above storage medium can be set to store a computer program for executing the following steps:

[0110] S201, receive input data from a multi-source heterogeneous data universe, where each data universe provides data with different dimensions and structures;

[0111] S202, convert the input data into a unified data lattice, and extract cross-dimensional feature data from the data lattice;

[0112] S203, perform data weaving processing on the cross-dimensional feature data, and output the data after the weaving processing in a hyper-dimensional form.

[0113] It can be seen that by receiving input data from a multi-source heterogeneous data universe, where each data universe provides data with different dimensions and structures; converting the input data into a unified data lattice, extracting cross-dimensional feature data from the data lattice; performing data weaving processing on the cross-dimensional feature data, and outputting the data after the weaving processing in a hyper-dimensional form, it is possible to have the flexibility to process different data types, while being able to extract effective information from complex data relationships, and generate a new data form, providing more in-depth data insights for various fields.

[0114] An embodiment of the present invention also provides an electronic device, including a memory and a processor, a computer program is stored in the memory, and the processor is set to run the computer program to execute the steps in any one of the above method embodiments.

[0115] Specifically, the above electronic device may further include a transmission device and an input / output device, where the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0116] Specifically, in this embodiment, the above-mentioned processor can be set to execute the following steps through a computer program:

[0117] S201, Receive input data from the multi-source heterogeneous data universe, where each data universe provides data with different dimensions and structures;

[0118] S202, Convert the input data into a unified data lattice, and extract cross-dimensional feature data from the data lattice;

[0119] S203, Perform data weaving processing on the cross-dimensional feature data, and output the data after weaving processing in a hyper-dimensional form.

[0120] Specifically, the specific examples in this embodiment can refer to the examples described in the above-mentioned embodiments and optional implementation manners, and will not be elaborated here.

[0121] It can be seen that by receiving input data from the multi-source heterogeneous data universe, where each data universe provides data with different dimensions and structures; converting the input data into a unified data lattice, and extracting cross-dimensional feature data from the data lattice; performing data weaving processing on the cross-dimensional feature data, and outputting the data after weaving processing in a hyper-dimensional form, it is possible to have the flexibility to process different data types, and at the same time be able to extract effective information from complex data relationships, and generate a new data form, providing deeper data insights for various fields.

[0122] The above has detailed the structure, features and effects of the present invention according to the embodiments shown in the drawings. The above is only the preferred embodiment of the present invention, but the present invention is not limited to the implementation scope shown in the drawings. Any changes made according to the concept of the present invention, or modified into equivalent embodiments with equivalent changes, still within the spirit covered by the specification and drawings, should be within the protection scope of the present invention.

Claims

1. A big data fusion calculation method, characterized in that, The method includes: Receiving input data from a multi-source heterogeneous data universe, where each data universe provides data with different dimensions and structures; Converting the input data into a unified data lattice, and extracting cross-dimensional feature data from the data lattice; wherein, converting the input data into a unified data lattice includes: regarding the input data as nodes in a graph, and using graph embedding technology to map the nodes into a multi-dimensional structure to obtain a multi-dimensional data lattice based on the graph structure, where the edges of the graph represent the relationships between the nodes; The extracting cross-dimensional feature data from the data lattice includes: converting each data point in the data lattice into an acoustic signal, combining the converted sound waves according to the positions of the data points in the data lattice, where each dimension can represent a sound channel; performing spectral analysis on the synthesized data sound, and using acoustic pattern recognition technology to extract the features of the data sound; using classification algorithms in machine learning to analyze the distribution of the features and classify the data; removing redundant features using principal component analysis, and using a feature selection algorithm to select the features most relevant to data classification; combining the acoustic features with other features in the original data to form a comprehensive data set containing the original data and cross-dimensional acoustic features, and outputting the integrated feature data set; Performing data weaving processing on the cross-dimensional feature data and outputting the woven data in a hyper-dimensional form.

2. The method according to claim 1, wherein The extracting cross-dimensional feature data from the data lattice includes: Performing persistent homology analysis on the multi-dimensional data lattice, and representing the analysis result of persistent homology as a topological summary of the data; Constructing a topological feature vector according to the topological summary, and combining topological feature vectors of different scales to generate a multi-dimensional feature representation.

3. The method according to claim 2, wherein The performing data weaving processing on the cross-dimensional feature data and outputting the woven data in a hyper-dimensional form includes: Mapping the cross-dimensional feature data into a pre-constructed symbol system, where each feature point or feature set in the cross-dimensional feature data is represented by one or a group of symbols; Performing spatial mapping on the cross-dimensional feature data to generate a hyper-dimensional representation space; Weaving the symbols in the symbol system in the hyper-dimensional representation space to generate and output a hyper-dimensional form.

4. A big data fusion computing system, characterized in that, The system includes: A receiving module for receiving input data from a multi-source heterogeneous data universe, where each data universe provides data with different dimensions and structures; A conversion module for converting the input data into a unified data lattice and extracting cross-dimensional feature data from the data lattice; wherein, converting the input data into a unified data lattice includes: regarding the input data as nodes in a graph, and using graph embedding technology to map the nodes into a multi-dimensional structure to obtain a multi-dimensional data lattice based on the graph structure, where the edges of the graph represent the relationships between the nodes; Extracting cross-dimensional feature data from the data lattice includes: converting each data point in the data lattice into an acoustic signal, combining the converted acoustic waves according to the positions of the data points in the data lattice, where each dimension can represent a sound channel; performing spectral analysis on the synthesized data sound, using acoustic pattern recognition technology to extract the features of the data sound; using classification algorithms in machine learning to analyze the distribution of the features and classify the data; adopting principal component analysis to remove redundant features, and using a feature selection algorithm to select the features most relevant to data classification; combining the acoustic features with other features in the original data to form a comprehensive data set containing the original data and cross-dimensional acoustic features, and outputting the integrated feature data set. A weaving module for performing data weaving processing on the cross-dimensional feature data and outputting the data after weaving processing in a hyperdimensional form.

5. The system according to claim 4, wherein The conversion module is specifically configured to: Perform persistent homology analysis on the multi-dimensional data lattice, and represent the analysis result of persistent homology as a topological summary of the data. Construct a topological feature vector according to the topological summary, and combine topological feature vectors of different scales to generate a multi-dimensional feature representation.

6. The system according to claim 5, wherein The weaving module is specifically configured to: Map the cross-dimensional feature data into a pre-constructed symbol system, where each feature point or feature set in the cross-dimensional feature data is represented by one or a group of symbols. Perform spatial mapping on the cross-dimensional feature data to generate a hyperdimensional representation space. Weave the symbols in the symbol system in the hyperdimensional representation space to generate and output a hyperdimensional form.

7. A storage medium, characterized in that, A computer program is stored in the storage medium, where the computer program is set to execute the method according to any one of claims 1-3 when running.

8. An electronic device, comprising a memory and a processor, characterized in that, A computer program is stored in the memory, and the processor is set to run the computer program to execute the method according to any one of claims 1-3.

Images