A multimodal data intelligent archiving management system

Abstract

This invention provides a multimodal data intelligent archiving and management system, relating to the field of multimodal data processing technology. It includes: a data mapping module, used to extract multimodal data from the native business environment, generate a data unit set, construct a virtual reference plane, establish a baseline rectangular grid topology and divide dynamic coordinate axes within the plane, map the data unit set to grid coordinates, and output the grid coordinate mapped data unit set; and an attribute tree construction module, used to construct a multidimensional attribute tree based on the density distribution of the grid coordinate mapped data unit set, generate dynamic evaluation nodes, calculate the weight parameter β value of each node, and output the attribute tree structure with β values. This invention achieves grid coordinate mapping, attribute tree weight evaluation, differentiated processing, blockchain notarization, hierarchical verification storage, and user behavior-driven dynamic optimization of multimodal data, thereby improving data independence, security, utilization efficiency, and optimizing management costs.

Description

Technical Field

[0001] This invention relates to the field of multimodal data processing technology, and in particular to a multimodal data intelligent archiving and management system. Background Technology

[0002] Traditional archiving technologies primarily employ a "format conversion + metadata annotation" processing model, which to some extent achieves archiving and storage of multimodal data. However, their ability to restore dynamic business relationships within the data is insufficient. Taking government approval systems as an example, the semantic relationship between process data (such as processing nodes and approval opinions) and supporting materials (such as scanned documents and on-site videos) relies on specific business logic parsing. While existing technologies can convert databases to common formats such as XML and annotate basic information through metadata, they struggle to fully preserve the complete semantic chain of implicit logical mapping relationships in complex business scenarios (such as the temporal relationship and semantic dependency between approval nodes and supporting materials).

[0003] The document mentions that "semantic extraction and resetting of structured discrete data is quite difficult." In practical applications, the historical data of a certain housing provident fund business system lacks a deep mapping of the dynamic semantics of business processes due to the use of traditional archiving technology, which makes it difficult to trace some business processes. This technical bottleneck makes the integrity of the business context of multimodal data easily affected after it is separated from the original system, thus restricting the in-depth utilization of the data. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide a multimodal intelligent data archiving and management system that enables trusted data storage throughout its entire lifecycle, cross-platform migration and utilization, and intelligent management and control.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] Firstly, a multimodal data intelligent archiving management system includes:

[0007] The data mapping module is used to extract multimodal data from the native business environment, generate data cell sets, construct a virtual reference plane, establish a baseline rectangular grid topology and divide dynamic coordinate axes in the plane, map the data cell sets to grid coordinates, and output the grid coordinate mapped data cell sets.

[0008] The attribute tree construction module is used to construct a multidimensional attribute tree based on the density distribution of the data cell set mapped by grid coordinates, generate dynamic evaluation nodes, calculate the weight parameter β value of each node, and output the attribute tree structure with β value.

[0009] The differentiation processing module is used to perform differentiation processing on the grid coordinate mapping data unit set according to the β value of each node and the preset hierarchical threshold, so as to obtain the processed data unit set.

[0010] The data packet generation module is used to convert the processed data unit set into a file set, and add a differentiated blockchain evidence identifier based on the attribute tree node β value, and output a pre-archived data packet with the β identifier;

[0011] The verification module is used to push data packets to the pre-archive verification end, perform hierarchical verification, output data packets that pass verification, parse and import the passed data packets into the database, allocate storage resource priorities, and output archived data with storage level identifiers.

[0012] The dynamic adjustment module provides retrieval services for archived data, collects user access behavior data in real time, feeds the access behavior data back to the attribute tree construction module, and dynamically adjusts the coordinate axis division rules.

[0013] Furthermore, multimodal data is extracted from the native business environment to generate a data cell set. A virtual reference plane is constructed, a baseline rectangular grid topology is established within the plane, and dynamic coordinate axes are defined. The data cell set is mapped to grid coordinates, and the grid coordinate mapped data cell set is output, including:

[0014] Extract multimodal data from the native business environment and generate independent data units carrying data type tags, confidentiality level tags, and call frequency tags;

[0015] Perform preprocessing operations on independent data units, including filtering valid data units based on preset rules, cleaning invalid fields, sorting sensitive fields according to confidentiality level requirements, and sorting data units according to call frequency, and outputting a set of preprocessed data units;

[0016] Using a preprocessed data unit set as input data, a three-dimensional virtual reference plane is constructed based on data type labels, security level labels, and frequency labels.

[0017] A reference rectangular grid topology covering the entire data range is established in a virtual reference plane, and dynamic coordinate axis scales are divided according to the distribution density of the input data in the three-dimensional feature space. Each preprocessed data unit is mapped to the corresponding coordinate unit of the grid, and the grid coordinate mapping data unit set is output.

[0018] Furthermore, a reference rectangular grid topology covering the entire data range is established within the virtual reference plane, and dynamic coordinate axis scales are divided according to the distribution density of the input data in the three-dimensional feature space. Each preprocessed data unit is mapped to the corresponding coordinate unit of the grid, outputting a set of grid coordinate mapped data units, including:

[0019] Based on the three-dimensional spatial coordinate range of the preprocessed data unit set, which is composed of data type, security level, and call frequency, a rectangular grid structure that completely covers all data coordinates is generated.

[0020] Based on the distribution density of each preprocessed data unit in three-dimensional space, the scale interval of each coordinate axis is adaptively adjusted, and each preprocessed data unit is located in the corresponding coordinate unit in the grid according to the data type label value, security level label value, and call frequency label value.

[0021] All data units that have completed coordinate positioning are integrated to form a grid coordinate mapping data unit set.

[0022] Furthermore, based on the density distribution of the grid coordinate mapping data cell set, a multidimensional attribute tree is constructed, dynamic evaluation nodes are generated, and the weight parameter β value of each node is calculated. The attribute tree structure with β values ​​is then output, including:

[0023] Based on the grid coordinate mapping data cell set, the number of data cells in each grid cell is counted as the density value;

[0024] A multidimensional attribute tree is constructed based on the number of data units in each grid cell as the density value. Each tree node corresponds to a grid cell, and the node level is determined by the position of the grid cell in the three-dimensional feature space.

[0025] For each tree node, a dynamic evaluation node is generated to obtain the density value of the corresponding grid cell. Based on the association relationship of three types of tags—data type, security level, and call frequency—the characteristic association strength value of the data cell within the corresponding grid cell is calculated.

[0026] Based on the density value of the corresponding grid cell and the characteristic association strength of the data cell within the corresponding grid cell, the node weight parameter β value is obtained, and a multidimensional attribute tree structure carrying the β value of each node is output.

[0027] Furthermore, based on the β value of each node and a preset hierarchical threshold, differential processing is performed on the grid coordinate mapping data cell set to obtain the processed data cell set, including:

[0028] Read the β values ​​of each node in the attribute tree structure with β values, and classify the node levels according to a preset threshold range, specifically including:

[0029] Nodes with a β value greater than or equal to the first preset threshold are defined as high β value nodes;

[0030] Nodes with a β value greater than or equal to the second preset threshold and less than the first preset threshold are defined as nodes with a mid-β value.

[0031] Nodes with a β value less than the second preset threshold are defined as low β value nodes;

[0032] Differentiated processing is performed on the grid coordinate mapping data unit set, including enhancing the sensitivity desensitization strength of data units associated with high β value nodes; increasing the data cleaning frequency of data units associated with medium β value nodes; and reducing the data sorting priority weight of data units associated with low β value nodes, outputting the optimized data unit set.

[0033] Furthermore, the processed data unit set is converted into a file set, and a differentiated blockchain evidence identifier is added based on the attribute tree node β value, outputting a pre-archived data package with the β identifier, including:

[0034] The optimized data unit set is converted into a structured file set;

[0035] Differentiated blockchain identifiers are generated based on the β values ​​of attribute tree nodes, specifically including:

[0036] Add a strong verification and evidence marker to the files corresponding to nodes with high β values;

[0037] Add a standard evidence identifier to the file corresponding to the β value node;

[0038] Add basic evidence identification to the files corresponding to nodes with low β values;

[0039] Each structured file is appended with a corresponding identifier, and a pre-archived data package with the β identifier is output.

[0040] Furthermore, the data packets are pushed to the pre-archiving verification end for hierarchical verification. Data packets that pass verification are output, and the passed data packets are parsed and imported into the database. Storage resource priorities are allocated, and archived data with storage level identifiers is output, including:

[0041] The pre-archived data packet with the β identifier is pushed to the pre-archived verification end. The corresponding strength of the verification process is triggered according to the β identifier level. If the verification passes, the parsing and import are performed to obtain the parsed data.

[0042] The parsed data is allocated storage resource priority based on the β identifier value, imported into the storage database, and the archived data is appended with a storage level identifier corresponding to the β value. The archived data with storage level identifiers is then output.

[0043] In a second aspect, a computing device includes:

[0044] One or more processors;

[0045] A storage device for storing one or more programs that, when executed by one or more processors, enable the one or more processors to implement the system.

[0046] Thirdly, a computer-readable storage medium storing a program that, when executed by a processor, implements the system.

[0047] The above-described solution of the present invention has at least the following beneficial effects:

[0048] Transforming multimodal data into a standardized grid coordinate system resolves the semantic gap between native data and archival systems, enabling data to be parsed independently of its native environment and improving integrity during cross-platform migration. Dynamic coordinate axes adaptively adjust based on data density, improving coordinate accuracy in high-frequency / high-sensitivity data areas and enhancing retrieval response speed. Beta values ​​are calculated by combining density distribution with tag association strength to quantify data business value and sensitivity. A multidimensional attribute tree structure explicitly reveals the data distribution logic, allowing business personnel to intuitively understand the sources of data importance (e.g., high density or high security level). High-beta data undergoes enhanced anonymization (e.g., full-field encryption), significantly reducing the risk of sensitive information leakage and meeting high-level security requirements; medium-beta data undergoes increased cleaning frequency, reducing integrity error rates. Low-beta data has lower sorting priority, freeing up computing resources and improving query speed for high-frequency data. Differentiated blockchain notarization identifiers are added based on beta values; high-value data is double-signed for notarization to ensure immutability, while notarization of medium- and low-value data is simplified to reduce on-chain storage costs. Standardized file set encapsulation resolves the heterogeneity of multimodal data formats and improves cross-system archiving compatibility.

[0049] High-beta data undergoes comprehensive and robust validation; low- and medium-beta data features simplified processes and improved overall validation efficiency. High-beta data is stored on SSDs, reducing access latency, while low-beta data is archived, effectively lowering costs and achieving "high-speed access to hot data and low-cost storage of cold data." Real-time collection of user access behavior automatically optimizes coordinate axis partitioning rules, continuously improving the retrieval hit rate of frequently queried data and enhancing the user experience. An adaptive learning mechanism allows data classification logic to evolve with business needs, reducing manual rule maintenance costs. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of a multimodal data intelligent archiving and management system provided by an embodiment of the present invention.

[0051] Figure 2 This is a flowchart illustrating the process of constructing a multidimensional attribute tree based on the density distribution of the data unit set mapped by grid coordinates, generating dynamic evaluation nodes, calculating the weight parameter β value of each node, and outputting the attribute tree structure with β value, according to an embodiment of the present invention. Detailed Implementation

[0052] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0053] like Figure 1 As shown, an embodiment of the present invention proposes a multimodal data intelligent archiving and management system, comprising:

[0054] The data mapping module is used to extract multimodal data from the native business environment, generate data cell sets, construct a virtual reference plane, establish a baseline rectangular grid topology and divide dynamic coordinate axes in the plane, map the data cell sets to grid coordinates, and output the grid coordinate mapped data cell sets.

[0055] The attribute tree construction module is used to construct a multidimensional attribute tree based on the density distribution of the data cell set mapped by grid coordinates, generate dynamic evaluation nodes, calculate the weight parameter β value of each node, and output the attribute tree structure with β value.

[0056] The differentiation processing module is used to perform differentiation processing on the grid coordinate mapping data unit set according to the β value of each node and the preset hierarchical threshold, so as to obtain the processed data unit set.

[0057] The data packet generation module is used to convert the processed data unit set into a file set, and add a differentiated blockchain evidence identifier based on the attribute tree node β value, and output a pre-archived data packet with the β identifier;

[0058] The verification module is used to push data packets to the pre-archive verification end, perform hierarchical verification, output data packets that pass verification, parse and import the passed data packets into the database, allocate storage resource priorities, and output archived data with storage level identifiers.

[0059] The dynamic adjustment module provides retrieval services for archived data, collects user access behavior data in real time, feeds the access behavior data back to the attribute tree construction module, and dynamically adjusts the coordinate axis division rules.

[0060] In this embodiment of the invention, a structured mapping of multimodal data from its native environment to standardized grid coordinates is achieved, enhancing the data's independent survival capability and cross-platform semantic consistency. A multidimensional attribute tree is constructed based on data density distribution, and weight β values ​​are calculated to provide a quantitative basis for data importance assessment and intelligent classification. Based on the β values, hierarchical desensitization, cleaning, and sorting optimization are performed on the data, improving the utilization efficiency of high-frequency data while ensuring the security of sensitive data. Differentiated evidence identifiers are added using blockchain technology to achieve tamper-proof evidence storage and operation traceability throughout the data's entire lifecycle. Through a hierarchical verification mechanism and dynamic allocation of storage resources, data quality is ensured and storage efficiency is optimized, with high-frequency, high-value data prioritized for allocation to high-speed storage media.

[0061] In a preferred embodiment of the present invention, multimodal data is extracted from the native business environment, a data unit set is generated, and a virtual reference plane is constructed. A baseline rectangular grid topology is established within the plane, and dynamic coordinate axes are divided. The data unit set is mapped to grid coordinates, and the grid coordinate-mapped data unit set is output. This process may include:

[0062] Extract multimodal data from the native business environment and generate independent data units carrying data type tags, confidentiality level tags, and call frequency tags;

[0063] Perform preprocessing operations on independent data units, including filtering valid data units based on preset rules, cleaning invalid fields, sorting sensitive fields according to confidentiality level requirements, and sorting data units according to call frequency, and outputting a set of preprocessed data units;

[0064] Using a preprocessed data unit set as input data, a three-dimensional virtual reference plane is constructed based on data type labels, security level labels, and frequency labels.

[0065] A baseline rectangular grid topology covering the entire data range is established within a virtual reference plane. Dynamic coordinate axis scales are then defined based on the distribution density of the input data in the three-dimensional feature space. Each preprocessed data cell is mapped to its corresponding grid coordinate cell, outputting a set of grid coordinate-mapped data cells, specifically including:

[0066] Based on the three-dimensional spatial coordinate range of the preprocessed data unit set, which is composed of data type, security level, and call frequency, a rectangular grid structure that completely covers all data coordinates is generated.

[0067] Based on the distribution density of each preprocessed data unit in three-dimensional space, the scale interval of each coordinate axis is adaptively adjusted, and each preprocessed data unit is located in the corresponding coordinate unit in the grid according to the data type label value, security level label value, and call frequency label value.

[0068] All data units that have completed coordinate positioning are integrated to form a grid coordinate mapping data unit set.

[0069] In this embodiment of the invention, multimodal data, including text, images, audio, and structured database tables, are collected from native business systems (such as government approval systems and financial transaction platforms). Full data capture is achieved through API interfaces or data interface adapters. For example, application form text, on-site inspection photos, approval voice recordings, and process database tables are extracted from government approval processes. Categories are defined based on data format (e.g., "text," "image," "structured data"), for example, scanning copies of ID cards are labeled "image," and approval opinion text is labeled "text." Security levels are classified according to business rules (e.g., public, internal, secret, confidential), for example, medical records involving personal privacy are labeled "secret," and publicly available policy documents are labeled "public." The number of times data is accessed within the past 12 months is statistically analyzed and categorized as high-frequency (greater than 100 times), medium-frequency (10-100 times), and low-frequency (less than 10 times). For example, frequently accessed user basic information tables are labeled "high-frequency," and low-frequency historical archive files are labeled "low-frequency."

[0070] Independent data unit preprocessing:

[0071] Invalid data is filtered based on preset rules (such as data integrity thresholds and business relevance indicators). For example, records missing key fields (such as empty ID numbers) are removed, or units exceeding the business validity period (such as temporary approval data from 3 years ago) are excluded. Outliers (such as garbled characters "###" in text fields) and duplicate values ​​(such as duplicate records in the same approval process) are identified and processed using string matching or clustering algorithms for standardization. For example, non-numeric characters in the "ID number" field are replaced with spaces. Differentiated anonymization is performed according to the confidentiality level.

[0072] Publicly available data: not anonymized (e.g., policy document titles);

[0073] Internal data: partially anonymized (e.g., replacing the middle 8 digits of the ID number with "****");

[0074] Confidential data: All fields are encrypted (e.g., medical diagnostic records are encrypted using the AES algorithm).

[0075] Data unit sorting: Sort by call frequency in descending order, with high-frequency data processed first. For example, frequently accessed user transaction records are placed at the front of the preprocessing queue, while low-frequency historical logs are placed at the back.

[0076] Construction of a 3D virtual reference plane:

[0077] X-axis: Data type (e.g., 0-100 range, where 0 represents text, 100 represents image, and the middle value represents mixed type);

[0078] Y-axis: Security level (0-100, 0 for public, 100 for confidential, linearly mapped according to security level weights).

[0079] Z-axis: Frequency of call (0-100, 0 is low frequency, 100 is high frequency, converted to coordinate values ​​using a logarithmic function to avoid high-frequency data being concentrated at the top).

[0080] The label value of each preprocessed data unit is normalized to the range [0, 100]. For example, an image file (type label value 80), secret level (confidentiality label value 70), and high frequency access (frequency label value 90) has three-dimensional coordinates of (80, 70, 90).

[0081] Baseline rectangular mesh topology and dynamic coordinate axis partitioning:

[0082] Calculate the maximum (MaxX, MaxY, MaxZ) and minimum (MinX, MinY, MinZ) values ​​of all data units in 3D space, and generate a minimum rectangular grid containing all coordinates, with a range of [MinX - Margin, MaxX + Margin] × [MinY - Margin, MaxY + Margin] × [MinZ - Margin, MaxZ + Margin]. Margin is an expansion threshold (e.g., 5% of the coordinate range) to ensure that edge data is covered.

[0083] The three-dimensional space is divided into an initial uniform grid (e.g., 10×10×10). The number of data units within each sub-grid is counted, and density is defined as: sub-grid data volume ÷ total data volume. For high-density regions (density greater than 1.5 times the mean), the scale interval is reduced (e.g., from 10 to 5), while for low-density regions (density less than 0.5 times the mean), the scale interval is increased (e.g., from 10 to 20). For example, high-frequency classified image data is concentrated in the region (70-90, 60-80, 80-100). The X, Y, and Z axis scale intervals in this region are adjusted from 10 to 5 to improve coordinate accuracy.

[0084] Data cell location and grid mapping:

[0085] Based on the adjusted dynamic scale, the normalized coordinates (X, Y, Z) of the data units are converted into grid coordinates (GridX, GridY, GridZ). For example, for a data unit with coordinates (85, 75, 95), in an area with an X-axis scale interval of 5, GridX = (85 - MinX) ÷ 5 (rounded down). The specific grid unit is located based on GridX, GridY, and GridZ. For example, the 3D coordinates (GridX=17, GridY=15, GridZ=19) correspond to the 17th row, 15th column, and 19th layer grid. If multiple data units are mapped to the same grid, they are sorted by hash value or timestamp to ensure unique record of each unit. All grid units are traversed to check for any missing coordinates or incorrect mappings. The mapping results are stored as a JSON or XML file with the structure "grid coordinates - data unit ID - label - original data pointer," forming a searchable mapping dataset. For example, each grid unit stores the index of all data units within that area, supporting quick lookup of corresponding data by coordinates.

[0086] By using tagging and grid mapping, heterogeneous data is uniformly transformed into a structured coordinate system, resolving the semantic gap between native data and the archival management system, enabling data to be parsed independently of the native system. The coordinate axis scale is automatically adjusted based on data density, improving coordinate accuracy in high-frequency / high-sensitivity data areas and reducing storage redundancy. Combining security level tags with grid positioning enables regionalized isolated storage of sensitive data (e.g., confidential data is concentrated in specific grid areas and encrypted separately), reducing the risk of data leakage. Three-dimensional coordinate mapping preserves the correlation between data type, access frequency, and security level. For example, the grid positioning of high-frequency confidential image data directly reflects its business importance, providing a basis for subsequent data utilization (e.g., priority archiving, access control). The standardized grid coordinate system supports lossless data migration between different business systems and hardware environments. For instance, when government data is migrated from an old approval system to a new archival management system, data associations can be quickly reconstructed through grid mapping.

[0087] In a preferred embodiment of the present invention, a multidimensional attribute tree is constructed based on the density distribution of the grid coordinate mapping data cell set, dynamic evaluation nodes are generated, and the weight parameter β value of each node is calculated. The attribute tree structure with β value is then output, which may include:

[0088] Based on the grid coordinate mapping data cell set, the number of data cells in each grid cell is counted as the density value;

[0089] A multidimensional attribute tree is constructed based on the number of data units in each grid cell as the density value. Each tree node corresponds to a grid cell, and the node level is determined by the position of the grid cell in the three-dimensional feature space.

[0090] For each tree node, a dynamic evaluation node is generated to obtain the density value of the corresponding grid cell. Based on the association relationship of three types of tags—data type, security level, and call frequency—the characteristic association strength value of the data cell within the corresponding grid cell is calculated.

[0091] Based on the density value of the corresponding grid cell and the characteristic association strength of the data cell within the corresponding grid cell, the node weight parameter β value is obtained, and a multidimensional attribute tree structure carrying the β value of each node is output.

[0092] In this embodiment of the invention, based on the generated set of grid coordinate mapping data units (such as a 100×100×100 three-dimensional grid), each grid unit (GridX, GridY, GridZ) is traversed in the order of X, Y, and Z axis coordinates. For example, starting from (0, 0, 0), all grids are visited row by row, column by column, and layer by layer. For each grid unit, the list of data unit IDs stored in that area is queried by index, and their quantity is counted as the density value. For example, if the grid (15, 20, 5) contains 200 data units, the density value is recorded as 200. If the grid unit is empty (density value = 0), it is marked as an "empty node" and treated as a leaf node when constructing the tree structure later. If the density value exceeds a threshold (such as twice the average), it is marked as a "high-density area" and child nodes are constructed first.

[0093] Hierarchical construction of multidimensional attribute trees:

[0094] The root node represents the entire 3D feature space, covering the coordinate range of all mesh cells (e.g., X∈[0, 100], Y∈[0, 100], Z∈[0, 100]). The first layer of child nodes divides the root node into several subspaces according to the X-axis coordinate (e.g., X<33, 33≤X<66, X≥66), with each subspace corresponding to a child node, and the node name recording the X-axis interval (e.g., "X0-33"). The second layer of child nodes further divides the first layer of child nodes according to the Y-axis coordinate (e.g., Y<33, 33≤Y<66, Y≥66), forming a two-dimensional subspace of "X interval - Y interval" (e.g., "X0-33_Y0-33"). The third layer of child nodes divides the Z-axis coordinate (e.g., Z<33, 33≤Z<66, Z≥66), finally forming a 3D subspace node (e.g., "X0-33_Y0-33_Z0-33"), with each node corresponding to a specific mesh cell.

[0095] Each tree node stores the range of grid coordinates it covers (e.g., X∈[a, b], Y∈[c, d], Z∈[e, f]) and associates it with a list of density values ​​for all grid cells within that range. For example, the third-level node “X0-33_Y0-33_Z0-33” corresponds to the first 33% of the region in the 3D grid, containing grid cells from (0, 0, 0) to (33, 33, 33).

[0096] Dynamic evaluation of node generation and characteristic correlation strength calculation:

[0097] For each tree node, the density values ​​of all grid cells it covers are aggregated, and a weighted average (e.g., weighted by grid area) is calculated as the node's base density index. For example, if a node covers 5 grid cells with density values ​​of 200, 150, 300, 100, and 250, its average density is (200+150+300+100+250)÷5=200. The distribution of data cell types within the node is statistically analyzed (e.g., 40% text, 50% images, 10% structured data), and the type diversity is calculated (e.g., using Shannon entropy: -0.4ln0.4 - 0.5ln0.5 - 0.1ln0.1). Higher diversity results in a stronger association. The distribution of security levels is analyzed (e.g., 30% public, 50% internal, 20% secret), and the proportion of high-security levels (secret and above) is calculated. A higher proportion results in a stronger association weight. The proportion of high-frequency data is statistically analyzed (e.g., 60% of data is accessed more than 100 times), and the high-frequency proportion is positively correlated with association strength. The indicators of the three types of labels are normalized (e.g., type diversity ∈ [0, 2], high density percentage ∈ [0, 1], high frequency percentage ∈ [0, 1]), and weight coefficients are set (e.g., type diversity weight 0.3, density weight 0.4, frequency weight 0.3); the characteristic association strength value is obtained by weighted summation = 0.3 × type entropy + 0.4 × high density percentage + 0.3 × high frequency percentage).

[0098] Calculation of node weight parameter β:

[0099] Normalize the node density values ​​to the global maximum density value to obtain a standardized density value ∈ [0, 1]. For example, if the global maximum density is 500 and a node density is 200, the standardized value is 200 ÷ 500 = 0.4. Normalize the characteristic association strength values ​​to the theoretical maximum value (e.g., assuming the maximum strength is 2) to obtain a standardized association strength ∈ [0, 1].

[0100] Comprehensive calculation of β value:

[0101] β = α × normalized density + (1 - α) × normalized association strength, where α is the density weighting coefficient (e.g., α = 0.6). If a tree node has child nodes, the β value can be obtained by weighted average of the β values ​​of the child nodes (e.g., weighted according to the number of grids covered by the child nodes), ensuring that the β value of high-level nodes in the tree structure reflects the overall importance.

[0102] Each node records the following information:

[0103] Spatial range (X / Y / Z interval);

[0104] Density value and standardized density;

[0105] Characteristic correlation strength value and normalized strength;

[0106] β value;

[0107] List of child nodes.

[0108] Convert the attribute tree to JSON or XML format for easy storage and transmission.

[0109] By comprehensively calculating density and the strength of the correlation between attributes, the β value can accurately reflect the business value and sensitivity of a data unit set. For example, data nodes with high density, high security level, and high frequency have higher β values ​​and are preferentially identified as core archive objects. The hierarchical structure of the attribute tree is dynamically generated according to the data distribution. High-density areas are automatically subdivided into more child nodes (such as high-frequency confidential data areas), realizing a differentiated strategy of "fine-grained management of important data and coarse-grained management of ordinary data," thereby improving storage and retrieval efficiency. The calculation of the strength of the correlation between attributes transforms the implicit correlation between data type, security level, and frequency into explicit values. For example, it discovers the pattern of "high-frequency correlation between image data and security level," providing a basis for optimizing data classification rules. The attribute tree displays the data distribution logic in a visual hierarchy. Business personnel can quickly understand the source of data importance through the node β value (e.g., whether a node's high β value is due to high density or high security level). It also supports adding new label dimensions (such as adding a "business type" label) to expand the tree structure.

[0110] In a preferred embodiment of the present invention, differential processing is performed on the grid coordinate mapping data unit set according to the β value of each node and a preset hierarchical threshold to obtain a processed data unit set, which may include:

[0111] Read the β values ​​of each node in the attribute tree structure with β values, and classify the node levels according to a preset threshold range, specifically including:

[0112] Nodes with a β value greater than or equal to the first preset threshold are defined as high β value nodes;

[0113] Nodes with a β value greater than or equal to the second preset threshold and less than the first preset threshold are defined as nodes with a mid-β value.

[0114] Nodes with a β value less than the second preset threshold are defined as low β value nodes;

[0115] Differentiated processing is performed on the grid coordinate mapping data unit set, including enhancing the sensitivity desensitization strength of data units associated with high β value nodes; increasing the data cleaning frequency of data units associated with medium β value nodes; and reducing the data sorting priority weight of data units associated with low β value nodes, outputting the optimized data unit set.

[0116] In an embodiment of the present invention, the root node is read from a stored attribute tree structure (such as in JSON or XML format), and all child nodes are recursively traversed to obtain the β value, spatial coordinate range, and associated data unit set index of each node. For example, starting from the root node "X0-100_Y0-100_Z0-100", child nodes such as "X0-33_Y0-33_Z0-33" are accessed in sequence, and the β value of each node (such as 0.8, 0.5, 0.2) is recorded. Each tree node is associated with a specific data unit set through the grid coordinate range. For example, the node "X66-100_Y66-100_Z66-100" corresponds to the last 34% of the area in the three-dimensional grid and is associated with a list of all data unit IDs in this area (such as ID-001, ID-002...).

[0117] Node level division and threshold matching:

[0118] The first preset threshold (Th1): such as 0.7 (threshold for high-β value nodes);

[0119] The second preset threshold (Th2): such as 0.3 (threshold for dividing medium and low-β value nodes).

[0120] High-β value nodes: β≥Th1 (such as β = 0.8≥0.7);

[0121] Medium-β value nodes: Th2≤β<Th1 (such as 0.3≤β = 0.5<0.7);

[0122] Low-β value nodes: β<Th2 (such as β = 0.2<0.3).

[0123] Enhanced desensitization intensity for data units of high-β value nodes:

[0124] According to the confidentiality level label of the data unit (such as "secret", "confidential"), the fields that need to be desensitized are located. For example, ID numbers, bank account numbers, medical diagnosis records, etc. in government affairs data.

[0125] Upgrading of desensitization rules:

[0126] Original desensitization rule: such as replacing the middle 8 digits of the ID number with " " (such as 110101 1234);

[0127] Enhanced rule: full-field encryption or replacement with an irreversible hash value (such as encrypted as "e10adc3949ba59abbe56e057f20f883e" through SHA-256).

[0128] Execution process:

[0129] Traverse the data units associated with nodes of high β values ​​(such as ID-001, ID-002);

[0130] Analyze the field structure of each data unit and identify sensitive fields;

[0131] Perform desensitization according to the enhanced rules, overwriting the original desensitization results;

[0132] Record the desensitization log (such as desensitization time and rule version).

[0133] The cleaning frequency of data cells in the mid-β value nodes increases:

[0134] The routine cleaning frequency is once a day, including filtering invalid values ​​(such as null values ​​and abnormal characters) and standardizing formats (such as unifying the date format to YYYY-MM-DD).

[0135] Frequency increased to 3 times per day:

[0136] Added three scheduled cleaning tasks in the morning, noon, and evening. Each cleaning task includes:

[0137] Duplicate value detection (such as duplicate submissions of the same approval form);

[0138] Cross-field logical validation (such as consistency checks between the "Age" field and the "Date of Birth" field);

[0139] Semantic error correction (such as verification of the checksum in the "ID number" field).

[0140] Incremental cleaning strategy:

[0141] Only data units updated since the last cleanup are cleaned, reducing resource consumption. For example, the third cleanup only processes data modified after 18:00 on the same day.

[0142] The sorting priority of data units with low β values ​​is reduced:

[0143] Sort by call frequency in descending order, with high-frequency data displayed first (e.g., the first 10 results in the query are high-frequency data).

[0144] Priority adjustment logic:

[0145] The sorting weight coefficient α is adjusted: the original α=1 (for high-frequency data), and the α for low β value data is reduced to 0.5; the new priority = call frequency × α + other factors (such as update time). For example, low β value data A: call frequency 100 times, original sorting score 100 × 1 = 100; after adjustment, score 100 × 0.5 = 50, and the sorting position moves from the 5th to the 15th position.

[0146] Validation and integration of the processed data unit set:

[0147] Randomly select 10% of high-β value data units and check whether the desensitization strength meets the enhancement rules (e.g., whether the full field encryption rate is 100%); check the cleaning logs of mid-β value data units to confirm whether the daily cleaning frequency reaches 3 times; verify whether the sorting results of low-β value data meet the weight reduction requirements. Reorganize the processed data unit set according to the original grid coordinate structure to generate a new dataset with processing labels (e.g., add a "processing level" field to the data unit metadata).

[0148] High-beta data nodes meet Level 3 requirements of the Cybersecurity Classified Protection Scheme 2.0 by enhanced data masking (e.g., full-field encryption). For example, the masking strength of "personal medical records" in government systems has been upgraded from partial replacement to irreversible encryption. For medium-beta data, increasing the frequency of data cleaning reduces the data integrity error rate from 5 to below 1. For example, the error rate of the "amount" field format in corporate financial data decreases. Low-beta data has its sorting priority reduced, improving the query response speed of high-frequency, high-value data (high-beta) by 40%, while reducing redundant computing resource consumption of low-value data (e.g., reducing CPU utilization by 15%). Processing strategies are dynamically adjusted based on beta value to achieve adaptive matching between "data importance" and "processing intensity." For example, quarterly financial report data (high-beta) automatically triggers the highest level of masking, while historical transaction logs (low-beta) only retain basic storage, reducing management costs.

[0149] In a preferred embodiment of the present invention, the processed data unit set is converted into a file set, and a differentiated blockchain evidence identifier is added based on the attribute tree node β value, outputting a pre-archived data package with the β identifier, which may include:

[0150] The optimized data unit set is converted into a structured file set;

[0151] Differentiated blockchain identifiers are generated based on the β values ​​of attribute tree nodes, specifically including:

[0152] Add a strong verification and evidence marker to the files corresponding to nodes with high β values;

[0153] Add a standard evidence identifier to the file corresponding to the β value node;

[0154] Add basic evidence identification to the files corresponding to nodes with low β values;

[0155] Each structured file is appended with a corresponding identifier, and a pre-archived data package with the β identifier is output.

[0156] In this embodiment of the invention, the processed data unit set is traversed, and the storage format is determined based on data tags (such as text, images, and structured tables).

[0157] Structured data (such as database tables): Convert to XML or JSON format, preserving field structure and relationships. For example, when converting government approval process data to XML, store it hierarchically according to "approval node → handler → attachments";

[0158] Semi-structured data (such as log files): Parsed into an array of JSON objects, key-value pairs retain the original fields (such as "timestamp: 2025-06-16" "operation type: query");

[0159] Unstructured data (such as images and audio): Keep the original binary format and attach metadata tags (such as file name, creation time, and data type).

[0160] File set organization rules:

[0161] Data is stored in folders based on grid coordinates. For example, the folder "X15_Y20_Z5" corresponds to the 3D coordinates (15,20,5), and contains all the data files for that area. Each file is named according to the following rule: "grid coordinates_data cell ID_β value.format suffix", such as "X15_Y20_Z5_ID001_0.85.xml".

[0162] Connect to a consortium blockchain or private blockchain network (such as a government blockchain platform), obtain node communication permissions, and configure consensus algorithms (such as PBFT) and encryption certificates.

[0163] Set evidence storage strategies corresponding to different β values:

[0164] High beta value (≥0.7): Invokes the smart contract to generate a strong verification identifier;

[0165] Medium β value (0.3~0.7): Generates standard evidence identification;

[0166] Low β value (<0.3): Generate basic evidence identifier.

[0167] Strong verification evidence generation (high β value node):

[0168] The file content is hashed using SHA-256 to obtain a fixed-length digest (e.g., a 64-bit hexadecimal string); the current UTC time (e.g., "2025-06-16T08:30:00Z") is obtained by calling the blockchain node API and concatenating it with the hash value; a signature string is generated by double-signing using the file creator's private key and the blockchain node's private key; the identifier structure is: {hash value, timestamp, signature string, proof level: "strong verification", β value: 0.85}.

[0169] Standard evidence storage identifier generation (mid-β value node):

[0170] Only SHA-256 hash + timestamp + single-node signature is executed, and the identifier structure is {hash value, timestamp, signature string, evidence level: "standard", β value: 0.5}.

[0171] Basic evidence storage identifier generation (low β value node):

[0172] Only the file hash value and blockchain block height are recorded. The identifier structure is {hash value, block height, evidence level: "basic", β value: 0.2}.

[0173] Differentiated identifiers appended to data packet encapsulation:

[0174] Structured / semi-structured files: Add an XML / JSON formatted evidence identification node to the file header, for example, insert it under the root node of an XML file. <blockchain-marker> ...< / blockchain-marker> Label;

[0175] Unstructured files: Generate independent marker files (.marker format), stored in the same path as the original file, such as "image.jpg" corresponding to "image.jpg.marker".

[0176] Pre-archived data packet encapsulation:

[0177] Package all files and identifiers according to the grid coordinate folder hierarchy using ZIP or TAR format; add a global index file (index.json) to the package file to record the mapping relationship of "grid coordinates - file list - β value - evidence level"; perform MD5 verification on the entire data package and generate a checksum to be appended to the package name (e.g., ...).

[0178] "archive_X0-33_Y0-33_Z0-33_20250616.md5").

[0179] Data packet integrity and compliance verification:

[0180] Parse the identifier of each file and query the blockchain explorer to see if the hash value is on the chain (e.g., input the hash value and verify the consistency between the block height and the timestamp); check if the signature string can be decrypted by the corresponding public key to ensure that it has not been tampered with. Check if the strong verification identifier of high beta files contains double signatures; check if the evidence storage identifier of sensitive data (such as ID card numbers) matches the desensitization level (e.g., fully encrypted data must correspond to a strong verification identifier).

[0181] Blockchain hash-based notarization ensures that any modification to high-beta data (such as core business records) results in a hash value change, making it tamper-proof once uploaded to the blockchain. High-beta data consumes more computing power for strong verification (such as double signatures), while low-beta data only records the basic hash, guaranteeing the strength of notarization for critical data. The notarization identifier includes elements such as timestamps and signatures, meeting the legal requirements for electronic archives. For example, during financial audits, the identifier can be used to quickly trace the creation and modification history of a transaction. The data package's index file is associated with the beta value, supporting rapid filtering of high-value data based on notarization level (strong verification / standard / basic). Regardless of whether the data is structured or unstructured, it is converted into a standardized file set and appended with a blockchain identifier, resolving the issue of heterogeneous native data formats and achieving cross-system and cross-platform archiving compatibility.

[0182] In a preferred embodiment of the present invention, pushing data packets to a pre-archiving verification end, performing hierarchical verification, outputting verified data packets, parsing and importing the verified data packets into the database, allocating storage resource priorities, and outputting archived data with storage level identifiers may include:

[0183] The pre-archived data packet with the β identifier is pushed to the pre-archived verification end. The corresponding strength of the verification process is triggered according to the β identifier level. If the verification passes, the parsing and import are performed to obtain the parsed data.

[0184] The parsed data is allocated storage resource priority based on the β identifier value, imported into the storage database, and the archived data is appended with a storage level identifier corresponding to the β value. The archived data with storage level identifiers is then output.

[0185] In this embodiment of the invention, pre-archived data packets (such as ZIP format) with a β identifier are pushed from the generation end to the pre-archived verification end server via FTP / SFTP or API interface. TLS encryption is enabled during transmission to ensure data integrity. For example, in a government data archiving system, high-β value data packets (such as sensitive social security data) are transmitted via a dedicated line, while low-β value data packets (such as publicly available policy documents) are transmitted via a regular network. After receiving the data packets, the verification end first reads the global index file (index.json) and parses the β value and evidence level (strong verification / standard / basic) of each file. For example, β=0.85 is extracted from the file "X15_Y20_Z5_ID001_0.85.xml", which is determined to be high-β value node data.

[0186] Pre-defined verification rule mapping:

[0187]

[0188] Strong validation execution details (high beta data):

[0189] Compare the data packet with the original generated checksum using MD5 checksum to ensure undamaged transmission; parse the blockchain evidence identifier and call the blockchain node API to verify the hash value, timestamp, and signature (e.g., use a public key to decrypt the signature string and compare the hash values ​​to see if they match); use an antivirus engine (such as Kaspersky) to scan the file for malicious code and check the anonymization strength of sensitive fields (e.g., whether confidential data is fully encrypted); compare the data unit label with the actual content, such as whether the "ID number" field conforms to the 18-digit rule and whether medical records contain required diagnostic fields.

[0190] Mid-level verification execution details (mid-level beta value data):

[0191] Deep virus detection in security scanning is omitted, and only a quick scan is performed. Semantic checks are simplified to field format validation (such as whether the date format is YYYY-MM-DD).

[0192] Details of weak validation execution (low beta data):

[0193] Only verify the MD5 hash of the data packet and the file format (such as whether XML has tag closure errors), skipping the deep verification of blockchain evidence.

[0194] The verification process generates a verification report (including verification time, β value, and passed items), such as "High β value data packet ID-20250616-001 passed verification, time taken 120 seconds"; a "verification passed" timestamp is added to the verified data packets to mark them as importable.

[0195] Handling verification failures:

[0196] Low β value data: Allow one automatic retry (e.g., retransmit the data packet), if it still fails, mark it as "pending manual processing";

[0197] For data with medium to high beta values: immediately trigger the manual review process, where the verifier manually checks the cause of the error (such as the signature being invalid due to the private key being expired), corrects it, and resubmits for verification.

[0198] Data packet parsing and database import:

[0199] For XML / JSON files, a parser (such as a DOM parser) is used to extract data fields and map them to database table structures. For example, the "applicant's name" field in a government approval XML file is mapped to the "applicant_name" database column; for unstructured files (such as images), thumbnails are generated and their paths are stored, while metadata (such as shooting time and resolution) is stored in database fields.

[0200] Batch import optimization:

[0201] High beta data is imported single records with real-time index updates to ensure immediate data availability.

[0202] For data with low to medium beta values, batch import is used (e.g., committing a transaction every 1000 records) to reduce database connection overhead.

[0203] Storage media mapping rules:

[0204] High β value (β≥0.7): allocated to SSD high-speed storage array, RAID level is RAID-10 (balancing speed and redundancy);

[0205] Medium β value (0.3≤β<0.7): Assigned to SAS hard drives, RAID-5;

[0206] Low β value (β<0.3): Assign to SATA hard drives or archive storage (such as tape libraries).

[0207] Storage class identifier generation:

[0208] Add a "storage_level" field to the database table with values ​​of "high / medium / low" corresponding to the β value range. For example, β=0.85 is marked as "high" and stored in the path " / ssd / storage / high / 2025 / 06 / ".

[0209] At the same time, update the index file to record the storage location and level.

[0210] The system monitors the utilization of storage at each level through a storage management system (such as Nagios). When the utilization rate of high-level storage exceeds 80%, an expansion warning is automatically triggered. Low-beta data is scanned periodically (e.g., monthly). If there are no access records within 6 months, the data is automatically migrated to offline archive storage, freeing up online storage resources. A robust verification process ensures the integrity and authenticity of highly sensitive data (such as bank transaction records), and, in conjunction with encrypted SSD storage, meets the Level 4 requirements of the Information Security Protection Standard 2.0. Medium- and low-sensitivity data employs appropriate verification and storage strategies to avoid resource waste caused by excessive protection. Storage level identifiers are linked to beta values, allowing operations personnel to quickly locate core business data through "high storage level," facilitating backup strategy formulation (e.g., daily incremental backups for high-level data, weekly full backups for low-level data). In case of failure, high-level data is prioritized for recovery, minimizing business interruption losses. The tiered storage architecture supports on-demand expansion of different levels of storage resources (e.g., adding SSD arrays to handle the growth of high-beta data), avoiding the waste of "full expansion" under traditional unified storage architectures.

[0211] Embodiments of the present invention also provide a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the system as described above. All implementations in the above system embodiments are applicable to this embodiment and can achieve the same technical effects.

[0212] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the system as described above. All implementations in the above system embodiments are applicable to this embodiment and can achieve the same technical effects.

[0213] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A multimodal data intelligent archiving management system, characterized in that, include: The data mapping module is used to extract multimodal data from the native business environment, generate data cell sets, construct a virtual reference plane, establish a baseline rectangular grid topology and divide dynamic coordinate axes in the plane, map the data cell sets to grid coordinates, and output the grid coordinate mapped data cell sets. The attribute tree construction module is used to construct a multidimensional attribute tree based on the density distribution of the data cell set mapped by grid coordinates, generate dynamic evaluation nodes, calculate the weight parameter β value of each node, and output the attribute tree structure with β value. The differentiation processing module is used to perform differentiation processing on the grid coordinate mapping data unit set according to the β value of each node and the preset hierarchical threshold, so as to obtain the processed data unit set. The data packet generation module is used to convert the processed data unit set into a file set, and add a differentiated blockchain evidence identifier based on the attribute tree node β value, and output a pre-archived data packet with the β identifier; The verification module is used to push data packets to the pre-archive verification end, perform hierarchical verification, output data packets that pass verification, parse and import the passed data packets into the database, allocate storage resource priorities, and output archived data with storage level identifiers. The dynamic adjustment module provides retrieval services for archived data, collects user access behavior data in real time, feeds the access behavior data back to the attribute tree construction module, and dynamically adjusts the coordinate axis division rules.

2. The multimodal data intelligent archiving management system according to claim 1, characterized in that, Multimodal data is extracted from the native business environment to generate a data cell set. A virtual reference plane is constructed, and a baseline rectangular grid topology is established within the plane, along with dynamic coordinate axes. The data cell set is then mapped to grid coordinates, and the grid coordinate-mapped data cell set is output, including: Extract multimodal data from the native business environment and generate independent data units carrying data type tags, confidentiality level tags, and call frequency tags; Perform preprocessing operations on independent data units, including filtering valid data units based on preset rules, cleaning invalid fields, sorting sensitive fields according to confidentiality level requirements, and sorting data units according to call frequency, and outputting a set of preprocessed data units; Using a preprocessed data unit set as input data, a three-dimensional virtual reference plane is constructed based on data type labels, security level labels, and frequency labels. A reference rectangular grid topology covering the entire data range is established in a virtual reference plane, and dynamic coordinate axis scales are divided according to the distribution density of the input data in the three-dimensional feature space. Each preprocessed data unit is mapped to the corresponding coordinate unit of the grid, and the grid coordinate mapping data unit set is output.

3. The multimodal data intelligent archiving management system according to claim 2, characterized in that, A baseline rectangular mesh topology covering the entire data range is established within a virtual reference plane. Dynamic coordinate axis scales are then defined based on the distribution density of the input data in the three-dimensional feature space. Each preprocessed data cell is mapped to its corresponding mesh coordinate cell, outputting a set of mesh coordinate-mapped data cells, including: Based on the three-dimensional spatial coordinate range of the preprocessed data unit set, which is composed of data type, security level, and call frequency, a rectangular grid structure that completely covers all data coordinates is generated. Based on the distribution density of each preprocessed data unit in three-dimensional space, the scale interval of each coordinate axis is adaptively adjusted, and each preprocessed data unit is located in the corresponding coordinate unit in the grid according to the data type label value, security level label value, and call frequency label value. All data units that have completed coordinate positioning are integrated to form a grid coordinate mapping data unit set.

4. The multimodal data intelligent archiving management system according to claim 3, characterized in that, Based on the density distribution of the grid coordinate mapping data cell set, a multidimensional attribute tree is constructed, dynamic evaluation nodes are generated, and the weight parameter β value of each node is calculated. The attribute tree structure with β values ​​is output, including: Based on the grid coordinate mapping data cell set, the number of data cells in each grid cell is counted as the density value; A multidimensional attribute tree is constructed based on the number of data units in each grid cell as the density value. Each tree node corresponds to a grid cell, and the node level is determined by the position of the grid cell in the three-dimensional feature space. For each tree node, a dynamic evaluation node is generated to obtain the density value of the corresponding grid cell. Based on the association relationship of three types of tags—data type, security level, and call frequency—the characteristic association strength value of the data cell within the corresponding grid cell is calculated. Based on the density value of the corresponding grid cell and the characteristic association strength of the data cell within the corresponding grid cell, the node weight parameter β value is obtained, and a multidimensional attribute tree structure carrying the β value of each node is output.

5. The multimodal data intelligent archiving management system according to claim 4, characterized in that, Based on the β value of each node and the preset hierarchical threshold, differential processing is performed on the grid coordinate mapping data cell set to obtain the processed data cell set, including: Read the β values ​​of each node in the attribute tree structure with β values, and classify the node levels according to a preset threshold range, specifically including: Nodes with a β value greater than or equal to the first preset threshold are defined as high β value nodes; Nodes with a β value greater than or equal to the second preset threshold and less than the first preset threshold are defined as nodes with a mid-β value. Nodes with a β value less than the second preset threshold are defined as low β value nodes; Differentiated processing is performed on the grid coordinate mapping data unit set, including enhancing the sensitivity desensitization strength of data units associated with high β value nodes; increasing the data cleaning frequency of data units associated with medium β value nodes; and reducing the data sorting priority weight of data units associated with low β value nodes, outputting the optimized data unit set.

6. The multimodal data intelligent archiving management system according to claim 5, characterized in that, The processed data unit set is converted into a file set, and a differentiated blockchain notarization identifier is added based on the attribute tree node β value. The output is a pre-archived data package with the β identifier, including: The optimized data unit set is converted into a structured file set; Differentiated blockchain identifiers are generated based on the β values ​​of attribute tree nodes, specifically including: Add a strong verification and evidence marker to the files corresponding to nodes with high β values; Add a standard evidence identifier to the file corresponding to the β value node; Add basic evidence identification to the files corresponding to nodes with low β values; Each structured file is appended with a corresponding identifier, and a pre-archived data package with the β identifier is output.

7. The multimodal data intelligent archiving management system according to claim 6, characterized in that, The data packets are pushed to the pre-archiving verification end, where hierarchical verification is performed. Data packets that pass verification are output, and the passed data packets are parsed and imported into the database. Storage resource priorities are allocated, and archived data with storage level identifiers is output, including: The pre-archived data packet with the β identifier is pushed to the pre-archived verification end. The corresponding strength of the verification process is triggered according to the β identifier level. If the verification passes, the parsing and import are performed to obtain the parsed data. The parsed data is allocated storage resource priority based on the β identifier value, imported into the storage database, and the archived data is appended with a storage level identifier corresponding to the β value. The archived data with storage level identifiers is then output.

8. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the system as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the system as described in any one of claims 1 to 7.

Images