A Method and System for Constructing a Natural Resource Business Data Association Pool Based on Spatiotemporal Correlation

Abstract

This invention provides a method and system for constructing a natural resource business data association pool based on spatiotemporal correlation. The method includes: extracting business data of various natural resource objects from multiple business systems responsible for managing each stage of natural resource management; cleaning the business data; generating fingerprint data of natural resource objects from the business data if data cleaning is complete; identifying the full-link trajectory of natural resource objects evolving at different times within the same space based on the fingerprint data; constructing an association pool based on the natural resource objects on the full-link trajectory; and performing business operations on the natural resource objects within the association pool based on the full-link trajectory. This embodiment overcomes the semantic barriers between different business systems, breaks down data silos, establishes correlations between business data, constructs a unified natural resource full-chain for natural resource objects, and traces and analyzes business data based on the topology of the association pool, forming full-cycle management and innovative applications.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of natural resource management, and particularly relates to a method and system for constructing a natural resource business data association pool based on space-time association. BACKGROUND

[0002] Natural resource management refers to the whole process of rational utilization, protection and sustainable development of natural resources. In the field of natural resource management, the whole life cycle management of the same natural resource object (such as a land plot) involves multiple links such as land use approval, land supply, reclamation and repair, and is supervised by different departments and carried by different business systems. The vertical management and operation of each business system lead to business data fragmentation and the formation of data islands.

[0003] Land use approval is a key means of use control. For the use control management department, a "use control system" is established as a land use compliance approval system. Land supply is a key activity of development and utilization. For the development and utilization management department, a "natural resource market" system is established to conduct process review and management of land transfer and allocation. Reclamation and repair is a key means of land consolidation. For the ecological restoration management department, a "land consolidation system" is established to manage the existing land reclamation and repair.

[0004] In the above systems, the "use control system" controls the land approval, the "natural resource market" controls the land supply, and the "land consolidation system" controls the land repair. From the perspective of a construction project, the approval of construction land approval, the contract of land supply, the approval of temporary land use, and the completion of temporary land reclamation are all in the physical world to promote the implementation of the construction project. The approval processes of land management links using different business systems are physically related.

[0005] In administrative management, on the one hand, there is a need to supervise the progress of the construction project, and on the other hand, there is a need to supervise the cross-business links such as approval without supply and supply without use. However, the current business system construction method will lead to insufficient business cooperation and supervision.

[0006] For example, in a use control management department, a "use control system" of independent business line is developed to support land use approval and management using mainstream computer and geographic information technology. The system adopts B / S architecture (browser / server), relies on distributed database storage space data (such as land plot boundary.shp file) and attribute data (such as land use scale, land use nature, index, etc.) in the back end, and connects graphic data, project (batch) information and text attachments and other data through case identification code, approval document number and other information.

[0007] In order to achieve online management of the land supply process, a "Natural Resources Market System" was built based on development frameworks such as Java and .Net, using a B / S architecture (browser / server). The business data of "maps, texts, and documents" within the land market are linked through contract numbers, electronic supervision numbers, etc., but it cannot form an accurate and effective connection with the data of the land use control system, or even any technical connection at all.

[0008] In the land consolidation system responsible for land reclamation and restoration, different software development frameworks, GIS underlying layers, and basic databases are selected. Its business data is internally transferred through land consolidation project numbers, while external business is manually connected through primitive means such as data extrapolation or table export.

[0009] For the same land parcel, its data is presented independently in different business systems:

[0010] In the land use control system, the land parcel exists in the form of a spatial layer and attribute table, containing data such as approval number, approved area, and approval time, but without land transfer or reclamation and restoration information.

[0011] The market supervision system records information such as the land transfer contract and the rights holder for the land parcel.

[0012] The land consolidation system preserves information on the initiation, implementation, and acceptance of land reclamation and restoration projects. Summary of the Invention

[0013] In view of this, the present invention provides a method and system for constructing a natural resource business data association pool based on spatiotemporal correlation, which is used to construct a unified natural resource full chain for natural resource objects.

[0014] The first aspect of the present invention provides a method for constructing a natural resource business data association pool based on spatiotemporal correlation, comprising:

[0015] Business data for various natural resource objects are retrieved from multiple business systems responsible for managing all aspects of natural resources.

[0016] Perform data cleaning on the aforementioned business data;

[0017] If data cleaning is completed, fingerprint data of the natural resource object is generated from the business data;

[0018] Based on the fingerprint data, the entire trajectory of the evolution of the natural resource object in the same space at different times can be identified;

[0019] Construct an association pool based on the natural resource objects along the entire trajectory;

[0020] In the association pool, business operations are performed on the natural resource object based on the full-link trajectory.

[0021] A second aspect of the present invention provides an apparatus for constructing a natural resource business data association pool based on spatiotemporal correlation, comprising:

[0022] The business data retrieval module is used to retrieve business data of various natural resource objects from multiple business systems responsible for the management of various aspects of natural resources.

[0023] The business data cleaning module is used to clean the business data.

[0024] The fingerprint data generation module is used to generate fingerprint data of the natural resource object from the business data if data cleaning is completed.

[0025] The end-to-end trajectory recognition module is used to identify the end-to-end trajectory of the natural resource object in the same space at different times based on the fingerprint data;

[0026] The association pool construction module is used to construct an association pool based on the natural resource objects on the entire link trajectory;

[0027] The business operation execution module is used to perform business operations on the natural resource object in the association pool according to the full-link trajectory.

[0028] A third aspect of the present invention provides a system for constructing a natural resource business data association pool based on spatiotemporal correlation, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the method for constructing a natural resource business data association pool based on spatiotemporal correlation as described in the first aspect above.

[0029] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for constructing a natural resource business data association pool based on spatiotemporal correlation as described in the first aspect above.

[0030] The fifth aspect of the present invention provides a computer program product that, when run on a computer, causes the computer to execute the method for constructing a natural resource business data association pool based on spatiotemporal correlation as described in the first aspect above.

[0031] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0032] In this embodiment, business data for various natural resource objects is retrieved from multiple business systems responsible for managing each stage of natural resource management; the business data is cleaned; if the data cleaning is complete, fingerprint data of natural resource objects is generated from the business data; based on the fingerprint data, the entire evolution trajectory of natural resource objects in the same space at different times is identified; an association pool is constructed based on the natural resource objects on the entire trajectory; and business operations are performed on the natural resource objects in the association pool according to the entire trajectory. This embodiment breaks through the semantic barriers between different business systems, overcomes data silos, establishes connections between business data, constructs a unified natural resource chain for natural resource objects, and traces and analyzes business data based on the topology of the association pool, forming full-cycle management and innovative applications. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 This is a schematic diagram of a method for constructing a natural resource business data association pool based on spatiotemporal correlation, provided by an embodiment of the present invention;

[0035] Figure 2 This is a system architecture diagram for processing business data provided in an embodiment of the present invention;

[0036] Figure 3 This is a topology diagram of a separately located structure provided in an embodiment of the present invention;

[0037] Figure 4 This is a topological structure diagram of urban and rural construction land in batches provided by an embodiment of the present invention;

[0038] Figure 5 This is a schematic diagram of a device for constructing a natural resource business data association pool based on spatiotemporal correlation, provided in an embodiment of the present invention.

[0039] Figure 6 This is a schematic diagram of a natural resource business data association pool construction system based on spatiotemporal correlation provided in an embodiment of the present invention. Detailed Implementation

[0040] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of the present invention. However, those skilled in the art will recognize that the present application may be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted to avoid unnecessary detail that could obscure the description of the present application.

[0041] The technical solution of the present invention will be illustrated below through specific embodiments.

[0042] The current situation of multiple departments having independent business systems and fragmented business data directly leads to the inability to release data value and inefficient system collaboration, becoming a key bottleneck restricting the management and intensive integration of the entire business process. The specific defects can be summarized as follows:

[0043] 1. Inconsistent encoding rules

[0044] The unique identification coding rules for land parcels differ among different business systems, and there is no technical mapping relationship, making it impossible to associate the same natural resource in different business systems through coding.

[0045] 2. Lack of data integration capabilities, resulting in data silos.

[0046] None of the business systems have reserved standardized data interaction interfaces, making it impossible to achieve real-time data synchronization.

[0047] The various business systems lack incremental data capture technology, making incremental data capture difficult and preventing real-time monitoring of data changes.

[0048] 3. Insufficient data traceability and analysis capabilities hinder full-cycle management and innovative applications.

[0049] The lack of records on the source, flow, and modification of data in each system results in a break in the data lineage, making it impossible to trace the entire lifecycle of land parcel data.

[0050] Because business data is scattered across different business systems, management decisions lack data support, making it impossible to conduct multi-dimensional correlation analysis across business systems and limiting the application of intelligent technologies such as AI (artificial intelligence) and big data.

[0051] This invention proposes a method and system for constructing a natural resource business data association pool based on spatiotemporal correlation. In accordance with the relevant requirements for further strengthening the collaborative processing and joint supervision of various related businesses, it coordinates the construction of common business service capabilities, clarifies the correlation between business data and the data collaboration and linkage between business systems, promotes the transformation of key business systems, promotes the full-process and full-cycle management of business data, and improves the intensive integration of business systems, business processing efficiency, and the level of innovation in business management models.

[0052] Reference Figure 1 The diagram illustrates a method for constructing a natural resource business data association pool based on spatiotemporal correlation, as provided in an embodiment of the present invention. Specifically, it may include the following steps:

[0053] Step 101: Extract business data of various natural resource objects from multiple business systems responsible for the management of various aspects of natural resources.

[0054] In this embodiment, ETL (Extract-Transform-Load) technology can be used to pull business data (such as approvals, transfer contracts (allocation decisions), recovery decisions, reclamation plans, etc.) of various natural resource objects from the data sources (such as databases, file systems, APIs (Application Programming Interfaces)) of business systems responsible for the management of various aspects of natural resources, and store the business data in the data pool.

[0055] The data pool is a centralized repository that stores a large amount of business data, including structured, semi-structured, and unstructured business data. It establishes relationships between various business data to form a relational pool, providing raw data support for subsequent processing and analysis.

[0056] Step 102: Clean the business data.

[0057] In this embodiment, by traversing the various business data in the data pool, one or more data cleaning processes can be performed on the business data to improve its quality.

[0058] In practice, basic data cleaning can be performed on business data, which mainly addresses issues of format and integrity.

[0059] Basic data cleaning includes at least one of the following: format standardization, filling in missing values, and deleting or correcting erroneous values.

[0060] For basic issues such as inconsistent field formats, missing values, and redundant values, batch processing can be performed using rule engines or script tools (such as Python Pandas or FME) as follows:

[0061] 1. Standardized format

[0062] 1.1 Numeric fields: Extract valid digits from character-type numeric fields using regular expressions and unify their units (e.g., convert "100 mu" to "6666.7 square meters").

[0063] 1.2. Date field: Match different date formats (such as "2023.10.01", "2023-10-01", "10 / 01 / 2023") through regular expressions and uniformly convert them to "YYYY-MM-DD".

[0064] 1.3. String field: Remove redundant spaces and unify the case (such as "CommercialLand" → "commercialland", and then map it to the Chinese "commercial land").

[0065] 2. Complete missing values

[0066] Derive and complete missing values based on business rules. For example, the administrative division code of the land parcel can be completed by matching the administrative division boundary according to the coordinates of the center point of the land parcel; the transfer period can be determined by the type of the land parcel (such as 40 years by default for commercial service land).

[0067] 3. Delete or correct error values

[0068] 3.1. Delete error values such as redundant data and invalid data. For example, area = -100 square meters, date = 9999-12-31, etc.

[0069] 3.2. Correct logical errors in error values. For example, the supply date is earlier than the approval date, and the repaired area is greater than the total area of the land parcel. etc. Abnormalities can be marked through the business rule library and verified and corrected manually.

[0070] If the basic data cleaning is completed, then spatial data cleaning is performed on the business data. Spatial data cleaning mainly solves the problem of the consistency between coordinates and graphics.

[0071] Among them, spatial data cleaning includes at least one of coordinate system conversion, graphic and attribute consistency verification, and spatial identifier association.

[0072] One of the core attributes of natural resource objects is spatial relevance. Spatial standardization is completed through professional GIS tools:

[0073] 1. Coordinate system conversion: Use coordinate conversion tools (such as the Gauss-Kruger projection conversion based on seven parameters) to uniformly convert the land parcel coordinates in the old coordinate system to the CGCS2000 coordinate system to ensure accurate alignment of spatial positions (the conversion error needs to be controlled within 0.5 meters to meet the management accuracy requirements of each link of natural resources).

[0074] 2. Graphic and attribute consistency verification: Calculate the actual area of the land parcel graphic (by calculating the vector graphic coordinates) and compare it with the registered area in the attribute table. If the error exceeds 5% (the threshold can be adjusted according to the business scenario), it is marked as abnormal and verified and corrected by the surveying and mapping department (such as re-measuring the land parcel boundary).

[0075] 3. Spatial Identifier Association: For business data that stores plot names and addresses but not coordinates, geocoding technology is used to convert text addresses (such as "No. 1, XX Road, XX District, XX City") into CGCS2000 coordinates. Conversely, for business data that stores plot names and coordinates but not addresses, geocoding technology is used to convert CGCS2000 coordinates into text addresses (such as "No. 1, XX Road, XX District, XX City"), thus associating the addresses and coordinates with a unified spatial framework.

[0076] This embodiment provides spatial data cleaning based on the characteristics of natural resource objects, building upon basic data cleaning, and gradually achieves unified data standards, integrating multi-source business data of natural resource objects to form structured business data.

[0077] Step 103: If data cleaning is completed, generate fingerprint data of natural resource objects from the business data.

[0078] When business data has been cleaned, natural language processing (NLP) techniques based on machine learning or deep learning can be used to fully understand the description of natural resource objects in each piece of business data, break through the semantic barriers between different business systems, and generate fingerprint data for each piece of business data.

[0079] Among them, the fingerprint data of natural resource objects are the characteristics that characterize natural resource objects and are used to identify natural resource objects that are unique in space.

[0080] This natural resource object serves as an anchor point for the fusion of multi-source business data. The fusion of multi-source business data is the foundation for realizing the full business process management of the natural resource object. It constructs an association pool covering multiple aspects such as land planning, basic geography, and natural resources, breaking down data silos and providing a foundation for subsequent analysis and application.

[0081] In one embodiment of the present invention, step 103 may include the following steps:

[0082] Step 1031: Label the attributes of each word group in the business data according to the word vector of each word group in the business data.

[0083] In this embodiment, the business data can be segmented into multiple word groups.

[0084] like Figure 2As shown, each word group in the business data is input into a language model such as BERT (Bidirectional Encoder Representations from Transformers) to learn the deep semantic representation of each word group, and thus obtain the word vector of each word group.

[0085] like Figure 2 As shown, the word vectors of each phrase in the business data are input into sequence labeling models such as CRF (Conditional Random Field) to label the attributes of each phrase, such as land parcel, land use, and usage period.

[0086] Step 1032: Filter out entity words representing natural resource objects and multiple contextual information of entity words from the phrases in the business data according to attributes.

[0087] In this embodiment, the attributes of each phrase can be traversed, and entity words representing natural resource objects (such as XX plot) and multiple contextual information of entity words (such as land location, land use, land area, term of use right, etc.) can be filtered from the phrases of business data according to the attributes.

[0088] Step 1033: Construct a graph structure using entity words and multiple contextual information.

[0089] In this embodiment, a structured undirected graph can be constructed using entity words and multiple contextual information, serving as a graph structure.

[0090] The graph structure includes nodes and edges. The graph structure can be represented as G=(V,E), where V is the set of graph nodes and E is the set of edges.

[0091] In a graph structure, nodes represent entity words and contextual information. Nodes in a graph structure include word vectors for entity words and word vectors for contextual information.

[0092] When there is a semantic relationship between entity words and context information (such as land parcel-location, land parcel-use), an edge is constructed between the node corresponding to the entity word and the node corresponding to the context information.

[0093] The initial weight of an edge can be set as the similarity between the features (initially word vectors) of the two nodes containing that edge.

[0094] like Figure 2 As shown, the features of nodes are updated through the message passing mechanism of GNN (Graph Neural Networks). When the update is completed, the weights between nodes are recorded as the degree of association between nodes. That is, the weights of edges in the graph structure represent the degree of association between entity words and various contextual information.

[0095] Among them, GNN is a neural network that runs directly on a graph to process the feature information of nodes, edges or the entire graph. It updates the features of the current node by aggregating the features of neighboring nodes, thereby capturing the dependencies and topological features between nodes in the graph.

[0096] Step 1034: Generate the first fingerprint fragment of the natural resource object based on the graph structure.

[0097] In this embodiment, feature engineering can be performed on the graph structure to convert it into a feature representation, denoted as the first fingerprint fragment of the natural resource object, where the dimension of the first fingerprint fragment is d_model.

[0098] In practical implementation, triplet data (head, relation, tail) can be extracted from the graph structure, where head is the head, relation is the relation, and tail is the tail. Both the head and tail are entity words with attributes and contextual information.

[0099] like Figure 2 As shown, triplet data is input into the TransR (R-type translation) model to generate the first fingerprint fragment of the natural resource object. The TransR model models entities and relations in two different spaces, namely the entity space and multiple relation spaces (relationship-specific entity spaces), and performs transformations in the corresponding relation spaces.

[0100] Step 1035: Generate a second fingerprint fragment of the natural resource object using an attention mechanism based on entity words and multiple contextual information.

[0101] For entity words and their contextual information, an attention mechanism can be used to construct a second fingerprint fragment of a natural resource object based on the semantics of the entity word itself and the semantics of multiple contextual information.

[0102] In one embodiment of the present invention, step 1035 may include the following steps:

[0103] Step 10351: Input the word vectors of entity words and word vectors of multiple contextual information into the multi-head attention model.

[0104] In this embodiment, as Figure 2 As shown, an MHA (Multi-Head Attention) model can be pre-trained, which provides an extended form of attention mechanism based on Transformer. The MHA model has multiple independent attention head structures, and obtains the attention distribution of the input in different subspaces by running multiple independent attention mechanisms in parallel, thereby capturing more comprehensive potential semantic associations in the input.

[0105] In this embodiment, the number of attention head structures is the same as the number of context information. Multiple attention head structures are input with word vectors of entity words, and one attention head structure is input with word vectors of one context information.

[0106] Step 10352: Guided by relevance, calculate the attention weight between the word vectors of entity words and the word vectors of context information in the attention head structure.

[0107] MHA has multiple attention heads, each responsible for processing the features of the land parcel and the features of the context information. Considering the correlation between the features of the land parcel and the features of the context information, two different processing schemes are configured for the attention heads. The attention weight between the word vectors of entity words and the word vectors of the context information is calculated in the attention head structure, which helps to improve robustness.

[0108] In the specific implementation, two processing schemes can be configured. For this, the average value of the correlation degree is calculated as the benchmark degree, and the benchmark degree is compared with the preset correlation threshold.

[0109] If the baseline is greater than or equal to the preset association threshold, then the first processing scheme is executed.

[0110] If the baseline is less than the preset correlation threshold, then the first processing scheme will be executed.

[0111] In the first processing scheme:

[0112] In the attention head structure, the word vectors of entity words in the attention mechanism are used as the query vector (i.e., Q vector), and the word vectors of context information are used as the key vector (i.e., K vector) and value vector (i.e. V vector) in the attention mechanism. The attention weight between the word vectors of entity words and the word vectors of context information is calculated according to the attention mechanism.

[0113] In this processing scheme, considering the significant differences in the descriptions of the same natural resource object across different business data and the presence of noise in the context information, the word vectors of entity words are used as the query target (i.e., Q vectors). The most relevant context information for the natural resource object is queried from all context information. The word vectors of the context information simultaneously provide the query basis (i.e., K vectors) and information to be weighted (i.e., V vectors), ensuring that the context information matching the natural resource object will be given priority.

[0114] The weight of contextual information is determined by its similarity to natural resource objects. The weight of contextual information with weak relevance to natural resource objects is reduced. Invalid contextual information is filtered out by entity words, thereby reducing the drift when describing natural resource objects between different business data.

[0115] In the second processing scheme:

[0116] In the attention head structure, the word vectors of entity words are used as the key vectors (i.e., K vectors) in the attention mechanism, and the word vectors of context information are used as the query vectors (i.e., Q vectors) and value vectors (i.e. V vectors) in the attention mechanism. The attention weights between the word vectors of entity words and the word vectors of context information are calculated according to the attention mechanism.

[0117] In this processing scheme, the word vector of the entity word is used as the reference standard (i.e., K vector). Each context information word vector actively queries its relevance to the entity word (i.e., Q vector). Then, the relevance is used to weight the context information word vector (i.e., V vector). Finally, the weighted result of all context information word vectors is anchored to the entity word vector, reducing the degree of deviation in description.

[0118] Each contextual information aligns with the entity word to obtain contribution information. In scenarios with multiple dimensions of contextual information (such as simultaneously including location, purpose, floor area ratio, and age), each dimension is strongly bound to the natural resource object, ensuring that all dimensions revolve around the natural resource object.

[0119] Step 10353: Fuse the word vectors of entity words with attention weights into neighborhood vectors.

[0120] In this embodiment, word vectors of entity words and corresponding attention weights can be fused into neighborhood vectors using methods such as concat and add.

[0121] Step 10354: Map the neighborhood vector to the second fingerprint fragment of the natural resource object.

[0122] In this embodiment, a fully connected layer (FC) or similar method can be used to map the neighborhood vector to a second fingerprint fragment of a natural resource object.

[0123] Step 1036: Merge the first fingerprint fragment and the second fingerprint fragment into fingerprint data of the natural resource object.

[0124] In this embodiment, as Figure 2 As shown, the first fingerprint fragment and the second fingerprint fragment can be merged into fingerprint data of a natural resource object using methods such as concat and add.

[0125] Step 104: Identify the full-link trajectory of the evolution of natural resource objects in the same space at different times based on fingerprint data.

[0126] In this embodiment, the concept of a full-link trajectory is abstracted for the management of each link of natural resources. The full-link trajectory is used to represent the evolution of natural resource objects in the same space at different times.

[0127] For different business data, the fingerprint data can be used to identify the entire trajectory of the evolution of natural resource objects in the same space at different times.

[0128] The entire chain trajectory includes projects, which refer to activities or projects involved by the natural resources department in exercising its responsibilities as the owner of all state-owned natural resource assets and in exercising its responsibilities for all land use control and ecological protection and restoration, such as engineering construction projects.

[0129] In some cases, the project also includes a certain type of business activity with a complete business management process in the business system, such as land use pre-approval, which mainly reflects the causes and effects of changes in natural resources.

[0130] Furthermore, the project is divided into main projects and sub-projects.

[0131] A master project refers to a super-class project created according to business rules to build the relationships throughout the entire project lifecycle. It is created at the beginning of the business process and serves as the master project to connect all sub-projects. It can be defined according to different application scenarios or themes.

[0132] Sub-projects are defined by conducting research and analysis on the project management processes of certain business systems and combining them with common capability building requirements to standardize the types of some sub-projects.

[0133] The main project can be divided into several stages according to time sequence, such as "pre-approval → approval → supply → use → replenishment → repair".

[0134] The meanings of each stage are as follows:

[0135] Pre-approval refers to the preliminary research and planning preparation stage. Before the natural resource object formally enters the "administrative approval process", the basic research, demonstration and plan preparation work is carried out around "whether it can be used, how to use it, and how much to use". It serves as the prerequisite for all subsequent stages.

[0136] "Approval (reporting for approval)" refers to the administrative approval and compliance confirmation process. It involves submitting the pre-approval plan to the competent authority for approval in accordance with legal authority and procedures, obtaining legal documents such as land use permits and planning permits, so that the use of natural resources has a legal basis.

[0137] Supply refers to the transfer and allocation of the right to use natural resources. After approval, the right to use natural resources is transferred from the owner to the user through legal means, clarifying who will use it, for how long, and under what conditions.

[0138] The use (utilization) refers to the actual development and dynamic supervision of a land parcel. After obtaining the land use right, the user carries out actual development, construction or production activities according to the approved purpose and plan, while the natural resources department conducts dynamic supervision on whether the land is used in accordance with regulations.

[0139] "Compensation" refers to the balancing and compensation process after resource occupation. When special protected resources (such as permanent basic farmland, arable land, forest land, and wetlands) are occupied by construction, development, or other activities, in order to maintain the total amount of resources and the quality of resources, the same type of resources with the same quantity and quality are re-allocated in other areas to achieve a balance between occupation and compensation.

[0140] Repair (restoration) refers to the process of restoring and improving the quality of ecology and land. It is a process that uses engineering, biological and other technical means to restore the ecological function or improve the quality of natural resource objects that have been damaged or degraded due to development and utilization, natural disasters (such as landslides and debris flows) or historical pollution (such as industrial site pollution) so that they can regain their utilization value (or return to ecological use).

[0141] Each stage has one or more sub-items. In different stages, some types of sub-items are mandatory, while others are optional.

[0142] During the pre-approval stage, sub-projects include land use pre-approval projects (mandatory), preliminary land use projects (optional), and so on.

[0143] During the approval phase, sub-projects include temporary land use projects (optional), separately located projects (mandatory), and so on.

[0144] During the supply phase, sub-projects include state-owned construction land transfer / allocation projects (mandatory), etc.

[0145] During the usage phase, sub-projects include development and utilization projects (mandatory), etc.

[0146] During the replenishment phase, sub-projects include replenishing arable land (optional), etc.

[0147] During the renovation phase, sub-projects include temporary land reclamation projects (optional), etc.

[0148] Some sub-projects have dependencies, with later sub-projects depending on earlier sub-projects.

[0149] In practical applications, from the perspective of engineering construction projects, obtaining the approval number for construction land, the contract number for land supply, and the approval number for temporary use, and completing the restoration and reclamation of the temporarily used land, are all aimed at promoting the implementation of the engineering construction project in the physical world. The approval process for each stage of land management is carried out through different business systems, and there are dependencies between different stages.

[0150] For example, see Figure 3 This shows a partial topology of the standalone site selection approval (sub-project) and other sub-projects; see alsoFigure 4 This shows a partial topology of urban and rural construction land in batches (sub-projects) and other sub-projects.

[0151] In a practical implementation, the similarity (such as cosine similarity) between the fingerprint data of two business data can be calculated, and the similarity can be compared with a preset similarity threshold.

[0152] If the similarity is greater than or equal to the preset similarity threshold, it indicates that the similarity is high. In this case, the two business data belong to the same space and are assigned to the same main project.

[0153] Under the main project, the word vectors of each phrase in the business data are concatenated with their corresponding fingerprint data to form a business enhancement vector.

[0154] like Figure 2 As shown, the business enhancement vector is input into a classifier such as MLP (Multilayer Perceptron), and the business data is classified into sub-items based on the business enhancement vector.

[0155] Generally, the classifier outputs the probability of business data belonging to each sub-item, and determines the sub-item with the highest probability of business data belonging to it.

[0156] like Figure 2 As shown, the business enhancement vector and the project vector of the sub-project are concatenated to form the target vector; where the project vector of the sub-project is a vector obtained by encoding the inherent data of the sub-project, such as its name and description.

[0157] The degree of dependency between two sub-items is calculated based on their target vectors using methods such as similarity and attention mechanisms.

[0158] In one approach, the distance between two sub-items can be calculated in the entire link trajectory as a relative position; where the distance is a logical distance, and the distance between two temporally adjacent sub-items is 1.

[0159] Since it is possible to statistically determine whether there is a dependency relationship between two sub-projects from historical main projects, thus forming a priori dependency relationship between the two sub-projects, such as a strong dependency between planning and construction, the existence of a dependency relationship between two sub-projects is not fixed in different main projects. The priori dependency relationship between the two sub-projects is initialized as an offset weight.

[0160] like Figure 2 As shown, in models that apply attention mechanisms, such as Transformer, the dependency between two sub-items is calculated according to the following formula:

[0161]

[0162] Where Attention represents the dependency level, Softmax is the activation function, Q represents the query vector, Q is the target vector of one of the sub-items, K represents the key vector, V is the value vector, K and V are both target vectors of another sub-item, T is the transpose matrix, and d k Let be the dimension, pos be the relative position, and w be the offset weight.

[0163] In this approach, relative temporal positions are added to models that employ attention mechanisms, such as Transformer, allowing Transformer to focus on recently relevant business and modulate based on prior knowledge, thereby improving the accuracy of dependency levels.

[0164] The dependency level is compared with the dependency threshold; the dependency relationship can be initially set to a default value.

[0165] If the dependency level is greater than or equal to the preset dependency threshold, it indicates that the dependency level between the two sub-projects is high, and a dependency relationship can be built between the two sub-projects.

[0166] Generally, the relationships between sub-projects can be established according to the time sequence of business data. The default relationship can be the basis for the dependency relationship. In this case, the dependency threshold with the largest value that meets the target condition can be found to correct the dependency relationship.

[0167] The objective condition is that dependencies cover associations.

[0168] Step 105: Construct an association pool based on natural resource objects along the entire link trajectory.

[0169] Within the data pool, a technical system and business hub for collaborative management are constructed. Natural resource objects are coded and indexed based on the entire chain trajectory. Then, businesses are connected with a single code to build a unified natural resource chain. This chain is dynamically updated as approvals from various business systems drive progress, forming an association pool.

[0170] Furthermore, coding refers to configuring identifiers for the main project and sub-projects based on a unified coding identifier system. The identifiers have the same spatiotemporal identifier and a time identifier. The identifiers are generated by fusing the time identifier and the spatial identifier.

[0171] Step 106: In the association pool, perform business operations on natural resource objects based on the full-link trajectory.

[0172] In this embodiment, data sources such as real-time monitoring streams and gridded structured data are integrated. By calling standardized APIs through change commands, synchronous updates of multiple business systems are achieved, incremental business data is captured, and multi-dimensional correlation analysis of natural resource objects across business systems can be carried out based on business data in the full-link trajectory in the association pool, enabling business operations using intelligent technologies such as AI and big data.

[0173] In the specific implementation, one of the business operations is access control. In this example, it can receive requests from the business system to query other sub-projects based on the current sub-project, that is, the events of other sub-projects are earlier than those of the current sub-project.

[0174] In response to the request, query the step size between the current sub-item and other sub-items in the full-link trajectory; where the step size is the logical distance, and the step size between two time-adjacent sub-items is 1.

[0175] The permission values ​​for the business system are calculated using the following formula:

[0176] Permission=e -α×max(step-β,γ) ×(δ×Dependency)

[0177] Where Permission is the permission value, e is a natural number, step is the step size, Dependency is the degree of dependency between the current sub-item and other sub-items, α, β, γ and δ are all hyperparameters (positive numbers), and max is a function that takes the maximum value.

[0178] If the permission value is greater than or equal to the preset permission threshold, the business data of other sub-projects in the association pool will be sent to the business system.

[0179] This includes adding the user identification information (such as department, name, mobile phone number, etc.) logged into the business system as a watermark to the business data of other sub-projects to improve the security of business data.

[0180] Suppose that under certain confidentiality requirements, users are allowed to trace back to the business data of sub-projects with a step size of 1. Under normal circumstances (i.e., the previous node and the next node are theoretically dependent), a high-privilege value is obtained.

[0181] When the step size is greater than 1, the permission value will drop rapidly. When the step size is close, if the dependency level is set to a high value, users can still be allowed to trace back to sub-items, realizing dynamic permission control and effectively improving the flexibility of permission control.

[0182] For example, the permission values ​​are as follows:

[0183] Permission=e -0.5×max(step-1,0)×(1.5×Dependency)

[0184] When applying different step sizes and levels of dependency, the permission values ​​can be as follows:

[0185] ;

[0186] In this example, the permission threshold is set to 0.45. Users can normally access sub-items with a step size of 1, as well as highly dependent sub-items with a step size of 2-3.

[0187] In this embodiment, business data for various natural resource objects is retrieved from multiple business systems responsible for managing each stage of natural resource management; the business data is cleaned; if the data cleaning is complete, fingerprint data of natural resource objects is generated from the business data; based on the fingerprint data, the entire evolution trajectory of natural resource objects in the same space at different times is identified; an association pool is constructed based on the natural resource objects on the entire trajectory; and business operations are performed on the natural resource objects in the association pool according to the entire trajectory. This embodiment breaks through the semantic barriers between different business systems, overcomes data silos, establishes connections between business data, constructs a unified natural resource chain for natural resource objects, and traces and analyzes business data based on the topology of the association pool, forming full-cycle management and innovative applications.

[0188] It should be noted that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0189] Reference Figure 5 The diagram illustrates a device for constructing a natural resource business data association pool based on spatiotemporal correlation, according to an embodiment of the present invention. Specifically, it may include the following modules:

[0190] The business data retrieval module 501 is used to retrieve business data of various natural resource objects from multiple business systems responsible for the management of various aspects of natural resources.

[0191] The business data cleaning module 502 is used to clean the business data.

[0192] The fingerprint data generation module 503 is used to generate fingerprint data of the natural resource object from the business data if data cleaning is completed.

[0193] The end-to-end trajectory recognition module 504 is used to recognize the end-to-end trajectory of the natural resource object in the same space at different times based on the fingerprint data.

[0194] The association pool construction module 505 is used to construct an association pool based on the natural resource objects on the full-link trajectory.

[0195] The business operation execution module 506 is used to perform business operations on the natural resource object in the association pool according to the full-link trajectory.

[0196] In one embodiment of the present invention, the business data cleaning module 502 includes:

[0197] The basic cleaning module is used to perform basic data cleaning on the business data; basic data cleaning includes at least one of format standardization, filling in missing values, and deleting or correcting erroneous values.

[0198] The spatial cleaning module is used to perform spatial data cleaning on the business data if basic data cleaning has been completed; spatial data cleaning includes at least one of coordinate system transformation, graphic and attribute consistency verification, and spatial identifier association.

[0199] In one embodiment of the present invention, the fingerprint data generation module 503 includes:

[0200] The attribute annotation module is used to annotate the attributes of each word group in the business data according to the word vector of each word group in the business data;

[0201] The phrase filtering module is used to filter out entity words representing the natural resource object and multiple contextual information of the entity words from phrases in the business data according to the attributes.

[0202] A graph structure construction module is used to construct a graph structure using the entity words and multiple contextual information.

[0203] The first fingerprint fragment generation module is used to generate a first fingerprint fragment of the natural resource object based on the graph structure.

[0204] The second fingerprint fragment generation module is used to generate a second fingerprint fragment of the natural resource object based on the entity words and multiple contextual information using an attention mechanism.

[0205] The fingerprint fragment fusion module is used to fuse the first fingerprint fragment and the second fingerprint fragment into fingerprint data of the natural resource object.

[0206] In one embodiment of the present invention, the nodes in the graph structure include word vectors of the entity words and word vectors of the context information, and the weights of the edges in the graph structure are the degree of association between the entity words and each piece of context information;

[0207] The second fingerprint fragment generation module includes:

[0208] A word vector input module is used to input the word vector of the entity word and the word vectors of multiple context information into a multi-head attention model; the multi-head attention model has multiple attention head structures, each of the multiple attention head structures inputs the word vector of the entity word, and each attention head structure inputs a word vector of the context information;

[0209] The attention weight calculation module is used to calculate the attention weight between the word vector of the entity word and the word vector of the context information in the attention head structure under the guidance of the relevance.

[0210] The neighborhood vector fusion module is used to fuse the word vectors of the entity words with the attention weights into a neighborhood vector;

[0211] The neighborhood vector mapping module is used to map the neighborhood vector to a second fingerprint fragment of the natural resource object.

[0212] In one embodiment of the present invention, the attention weight calculation module includes:

[0213] The benchmark calculation module is used to calculate the average value of the correlation degree as the benchmark degree.

[0214] The first weight calculation module is used to calculate the attention weight between the word vector of the entity word and the word vector of the context information in the attention head structure if the benchmark degree is greater than or equal to a preset association threshold.

[0215] The second weight calculation module is used to calculate the attention weight between the word vector of the entity word and the word vector of the context information in the attention head structure if the benchmark degree is less than a preset association threshold.

[0216] In one embodiment of the present invention, the end-to-end trajectory includes a main project, which is divided into multiple stages in chronological order, and each stage has one or more sub-projects;

[0217] The end-to-end trajectory recognition module 504 includes:

[0218] The similarity calculation module is used to calculate the similarity between the fingerprint data of two pieces of business data;

[0219] The main project division module is used to determine that two pieces of business data belong to the same space if the similarity is greater than or equal to a preset similarity threshold, and to divide the two pieces of business data into the same main project;

[0220] The business enhancement vector concatenation module is used to concatenate the word vectors of each word group in the business data with the fingerprint data to form a business enhancement vector under the main project.

[0221] The sub-item classification module is used to classify the sub-items to which the business data belongs based on the business enhancement vector;

[0222] The target vector concatenation module is used to concatenate the business enhancement vector with the project vector of the sub-project to form a target vector.

[0223] The dependency calculation module is used to calculate the dependency between the two sub-projects based on their target vectors.

[0224] A dependency building module is used to build a dependency relationship between two sub-projects if the dependency level is greater than or equal to a dependency threshold.

[0225] The dependency correction module is used to find the dependency threshold that satisfies the target condition and has the largest value, so as to correct the dependency relationship; wherein, the target condition is that the dependency relationship covers the association relationship established between the sub-projects according to the time order between the business data.

[0226] In one embodiment of the present invention, the dependency calculation module includes:

[0227] The relative position statistics module is used to calculate the distance between two sub-items in the full-link trajectory as their relative position.

[0228] The offset weight initialization module is used to initialize the prior dependency relationship between the two sub-items as offset weights;

[0229] The position offset calculation module is used to calculate the dependency between the two sub-items according to the following formula: Where Attention represents the degree of dependency, Softmax is the activation function, Q represents the query vector, Q is the target vector of one of the sub-items, K represents the key vector, V represents the value vector, K and V are both target vectors of another sub-item, T is the transpose matrix, and d k Let be the dimension, pos be the relative position, and w be the offset weight.

[0230] In one embodiment of the present invention, the business operation execution module 506 includes:

[0231] The query request receiving module is used to receive requests from the business system to query other sub-projects based on the current sub-project.

[0232] The step size query module is used to query the step size of the interval between the current sub-item and other sub-items in the full-link trajectory;

[0233] The permission value calculation module is used to calculate the permission value of the business system according to the following formula: Permission = e -α×max(step-β,γ) ×(δ×Dependency); where Permission is the permission value, e is a natural number, step is the step size, Dependency is the degree of dependency between the current sub-item and other sub-items, α, β, γ and δ are all hyperparameters, and max is a function that takes the maximum value;

[0234] The business data sending module is used to send the business data of other sub-items in the association pool to the business system if the permission value is greater than or equal to a preset permission threshold.

[0235] This invention provides a device for constructing a natural resource business data association pool based on spatiotemporal correlation. By applying this device, the various steps in the aforementioned embodiments of the methods for constructing natural resource business data association pools based on spatiotemporal correlation can be implemented.

[0236] It should be noted that the module division in the various spatiotemporal correlation-based natural resource business data association pool construction devices provided in the above embodiments is illustrative and only represents a logical functional division. In actual implementation, other division methods may also be used. Furthermore, the functional modules in the various embodiments of this invention can be integrated into a single processor, exist as separate physical entities, or be integrated into a single module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0237] If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the technical solution of the embodiments of the present invention can be embodied in the form of a computer program product, which is stored in a computer storage medium and includes several instructions to cause a spatiotemporal correlation-based natural resource business data association pool construction system or processor to execute all or part of the steps of the methods in the various embodiments of the present invention. The aforementioned computer storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0238] Furthermore, the apparatus for constructing a natural resource business data association pool based on spatiotemporal correlation provided in the above embodiments and the method for constructing a natural resource business data association pool based on spatiotemporal correlation belong to the same concept. For details of their specific implementation process, please refer to the method embodiments, which will not be repeated here.

[0239] Reference Figure 6 This diagram illustrates a system for constructing a natural resource business data association pool based on spatiotemporal correlation, as provided in an embodiment of the present invention. Figure 6 As shown, the spatiotemporal correlation-based natural resource business data association pool construction system in this embodiment of the invention includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps in the above-described spatiotemporal correlation-based natural resource business data association pool construction method embodiment. Alternatively, when the processor executes the computer program, it implements the functions of each module in the above-described spatiotemporal correlation-based natural resource business data association pool construction device embodiment.

[0240] For example, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to complete this application. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which can be used to describe the execution process of the computer program in the spatiotemporal correlation-based natural resource business data association pool construction system.

[0241] The system for constructing a natural resource business data association pool based on spatiotemporal correlation can be an electronic device such as a desktop computer or a cloud server. This system may include, but is not limited to, processors and memory. Those skilled in the art will understand that... Figure 6 This is merely one example of a system for constructing a natural resource business data association pool based on spatiotemporal correlation, and does not constitute a limitation on such a system. It may include more or fewer components than illustrated, or combine certain components, or use different components. For example, the system may also include input / output devices, network access devices, buses, etc.

[0242] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0243] The memory can be an internal storage unit of the spatiotemporal correlation-based natural resource business data association pool construction system, such as a hard drive or memory of the system. Alternatively, the memory can be an external storage device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc., equipped on the system. Furthermore, the memory can include both internal storage units and external storage devices. The memory is used to store the computer program and other programs and data required by the spatiotemporal correlation-based natural resource business data association pool construction system. The memory can also be used to temporarily store data that has been output or will be output.

[0244] This invention also discloses a system for constructing a natural resource business data association pool based on spatiotemporal correlation, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the method for constructing a natural resource business data association pool based on spatiotemporal correlation as described in the foregoing embodiments.

[0245] This invention also discloses a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method for constructing a natural resource business data association pool based on spatiotemporal correlation as described in the foregoing embodiments.

[0246] This invention also discloses a computer program product that, when run on a computer, causes the computer to execute the method for constructing a natural resource business data association pool based on spatiotemporal correlation as described in the foregoing embodiments.

[0247] The embodiments described above are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for constructing a natural resource business data association pool based on spatio-temporal association, characterized in that, The method comprises the following steps: pulled from a plurality of business systems responsible for the management of various aspects of natural resources, the business data of various natural resource objects; data cleaning is performed on the business data; if the data cleaning is completed, the attributes of each phrase in the business data are labeled according to the word vectors of each phrase in the business data; filtering out entity words representing the natural resource objects and a plurality of context information of the entity words from the phrases in the business data according to the attributes; constructing a graph structure using the entity words and a plurality of the context information; generating a first fingerprint segment of the natural resource object according to the graph structure; inputting the word vectors of the entity words and the word vectors of a plurality of the context information into a multi-head attention model; the multi-head attention model has a plurality of attention head structures, and the word vectors of the entity words are input into each of the plurality of attention head structures, and the word vectors of one of the context information are input into one of the attention head structures; calculating the average of the correlation degree between the entity words and each of the context information as a reference degree; if the reference degree is greater than or equal to a preset correlation threshold, then in the attention head structure, the word vector of the entity word is used as the query vector, and the word vector of the context information is used as the key vector and the value vector, and the attention weight between the word vector of the entity word and the word vector of the context information is calculated according to the attention mechanism; if the reference degree is less than the preset correlation threshold, then in the attention head structure, the word vector of the entity word is used as the key vector, and the word vector of the context information is used as the query vector and the value vector, and the attention weight between the word vector of the entity word and the word vector of the context information is calculated according to the attention mechanism; fusing the word vector of the entity word and the attention weight into a neighborhood vector; mapping the neighborhood vector into a second fingerprint segment of the natural resource object; fusing the first fingerprint segment and the second fingerprint segment into fingerprint data of the natural resource object; identifying the full-link trajectory of the natural resource object evolving at different times in the same space according to the fingerprint data; constructing a correlation pool according to the natural resource object on the full-link trajectory; performing business operations on the natural resource object in the correlation pool according to the full-link trajectory.

2. The method of claim 1, wherein, The data cleaning of the business data comprises: performing basic data cleaning on the business data; the basic data cleaning includes at least one of format standardization, missing value completion, deletion or correction of error values; if the basic data cleaning is completed, then performing spatial data cleaning on the business data; the spatial data cleaning includes at least one of coordinate system conversion, consistency verification of graphics and attributes, and spatial identification association.

3. The method of claim 1, wherein, The nodes in the graph structure include the word vectors of the entity words and the word vectors of the context information, and the weights of the edges in the graph structure are the correlation degrees between the entity words and each of the context information.

4. The method according to any one of claims 1 to 3, characterized in that, The full-link trajectory includes a main project, a plurality of stages are divided in time sequence under the main project, and each of the stages has one or more sub-projects; The full-link trajectory of evolution of the natural resource object in different times in the same space according to the fingerprint data comprises: calculating the similarity between the fingerprint data of two pieces of business data; if the similarity is greater than or equal to a preset similarity threshold, determining that the two pieces of business data belong to the same space, and dividing the two pieces of business data into the same main project; concatenating the word vectors of each phrase in the business data with the fingerprint data as a business-enhanced vector under the main project; classifying the sub-projects to which the business data belongs according to the business-enhanced vector; concatenating the business-enhanced vector with a project vector of the sub-project as a target vector; calculating the dependency degree between two sub-projects according to the target vectors of the two sub-projects; if the dependency degree is greater than or equal to a dependency threshold, constructing a dependency relationship between the two sub-projects; finding the dependency threshold that satisfies the target condition and has the largest value, to correct the dependency relationship; wherein the target condition is that the dependency relationship covers the association relationship between the sub-projects established according to the time sequence of the business data.

5. The method of claim 4, wherein, The method for calculating the dependency degree between two sub-projects according to the target vectors of the two sub-projects comprises: statistically calculating the distance between two sub-projects in the full-link trajectory as a relative position; initializing the prior dependency relationship between the two sub-projects as a bias weight. The dependency degree between two of the sub-items is calculated according to the following formula: ; wherein, Attention is the dependency degree, Softmax is an activation function, Q represents a query vector, Q is a target vector of one of the sub-items, K represents a key vector, V represents a value vector, K and V are target vectors of another of the sub-items, T is a transposed matrix, d k is a dimension, and pos is the relative position, and w is the offset weight.

6. The method of claim 4, wherein, The method for performing a business operation on the natural resource object according to the full-link trajectory in the association pool comprises: receiving a request of the business system for querying other sub-projects based on the current sub-project; querying the step length between the current sub-project and other sub-projects in the full-link trajectory; A permission value of the business system is calculated according to the following formula: Permission = e -α×max(step-β,γ) × (δ × Dependency); wherein, Permission is the permission value, e is a natural number, step is the step, Dependency is the dependency degree between the current sub-project and other sub-projects, and α, β, γ and δ are all hyperparameters, and max is a function of taking the maximum value. if the permission value is greater than or equal to a preset permission threshold, sending the business data of other sub-projects in the association pool to the business system.

7. A spatio-temporal correlation based natural resource business data correlation pool construction system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program to realize the method for constructing an association pool of natural resource business data based on spatio-temporal association according to any one of claims 1-6.

8. A computer-readable storage medium storing a computer program, the computer-readable storage medium comprising: The computer program is executed by the processor to realize the method for constructing an association pool of natural resource business data based on spatio-temporal association according to any one of claims 1-6.

Images