Land reserve data multi-source integration and dynamic updating method and system

Abstract

The invention relates to the technical field of land reserve, in particular to a multi-source integration and dynamic updating method and system for land reserve data, and the method comprises the steps: obtaining multi-source land data, constructing a unified space-time database through hierarchical coordinate fusion, and achieving the data integration. And then, land parcel ownership characteristics are extracted, compliance evaluation is completed, and policy change data are synchronously analyzed to generate planning constraints. Idle risk assessment is carried out based on spatio-temporal data and policy constraints, land value fluctuation is predicted, and accurate valuation data is obtained through idle risk weighted correction. And finally, according to risk assessment and policy constraints, a low-efficiency land taking strategy is generated, all data and strategies are uploaded to a management platform in real time, value abnormal block marking and land reserve dynamic optimization tasks are executed respectively, and full-process closed-loop management from data integration to intelligent decision making is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of land reserve, and particularly relates to a land reserve data multi-source integration and dynamic updating method and system. BACKGROUND

[0002] With the deepening of urbanization process, land resources as a scarce public asset, its efficient management and optimal allocation has become the core issue of sustainable development of city. Land reserve is the key means for government to implement macro-control, guarantee key project landing and promote land intensive and economical use. However, the existing land reserve management mode faces significant challenges in data integration and dynamic updating.

[0003] Firstly, the data relied on by land reserve management has typical multi-source heterogeneous characteristics, not only including government data such as land survey, real estate registration and planning permission, but also involving multi-dimensional information such as market transaction and environmental assessment. These data have different sources, various formats and different space-time benchmarks, and are independently managed by different departments, forming a serious information island. Secondly, the state of land reserve is not static, but is dynamically affected by external factors such as policy and regulation, market situation and planning adjustment. For example, new urban planning may adjust the land use, and macroeconomic policy may affect the land value. These changes require that the reserve information must be updated in real time or near real time. The existing method lacks an effective mechanism to actively capture and analyze these external changing factors, and quantitatively feedback their influence to the reserve database, so that the management strategy is often lagging behind the actual situation, and it is difficult to realize the transition from static management to dynamic supervision. Finally, the conversion chain from data to decision is broken. The current system is mostly limited to data storage and display, and cannot deeply mine data correlation, cannot make forward-looking prediction on land idle risk, and lacks intelligent research and judgment on land asset value fluctuation. At the same time, for the identified problem land, such as potential idle or inefficient use land, it lacks the ability of automatic strategy generation based on data driving, and it is difficult to provide accurate and efficient decision support for revitalizing the stock resources and optimizing the reserve structure.

[0004] The above content is only used to assist in understanding the technical solutions of the present application, and does not represent the acknowledgement of the above content as prior art. SUMMARY

[0005] The main purpose of the present application is to provide a land reserve data multi-source integration and dynamic updating method and system, which aims to solve the technical problems that the existing land reserve management has difficulty in multi-source heterogeneous data integration, dynamic policy influence analysis lags behind, idle risk and value evaluation lacks intelligence, and the revitalization strategy lacks data driving, resulting in insufficient scientificity and timeliness of land reserve management.

[0006] To achieve the above objectives, the present invention provides a method for multi-source integration and dynamic updating of land reserve data, the method comprising: Acquire multi-source data of land reserve areas and perform hierarchical data coordinate fusion on the multi-source data of land reserve areas to obtain a unified spatiotemporal database of land reserves; Land ownership characteristics are extracted from the unified spatiotemporal database of land reserves to obtain land ownership data, and land compliance assessment is conducted on the land ownership data to obtain land compliance assessment data. Obtain policy change data for land reserve areas, and perform dynamic analysis of planning constraints based on this data to obtain policy constraint data; conduct regional land idling risk assessment based on the unified spatiotemporal database of land reserves and policy constraint data to obtain regional land idling risk data. Land value fluctuations are predicted based on land compliance assessment data to obtain land value prediction data; the land value prediction data is then adjusted for the impact of idle land based on regional land idle risk data to obtain corrected land value data, which is then uploaded to the land reserve management platform to perform the task of marking abnormal value blocks. Based on regional land idling risk data and policy constraint data, strategies for revitalizing inefficient land use are generated, and these strategies are uploaded to the land reserve management platform to execute dynamic optimization tasks for land reserves.

[0007] Optionally, the step of acquiring multi-source data of the land reserve area and performing hierarchical data coordinate fusion on the multi-source data of the land reserve area to obtain a unified spatiotemporal database of land reserves includes: Obtain multi-source data on land reserve areas, including geographic information system data, real estate registration data, urban planning map data, and land market listing data; Data timeliness verification and coordinate system calibration are performed on multi-source data of land reserve areas to obtain a standardized dataset; Land reserve tiers are divided based on standardized datasets to obtain primary and secondary reserve data. Based on a unified spatiotemporal benchmark, data from primary and secondary reserve warehouses are dynamically correlated to obtain time-series correlation data for reserves. By integrating the ownership boundaries of the land reserve time-series data, a unified spatiotemporal database of land reserves can be obtained.

[0008] Optionally, the step of extracting land ownership characteristics from the unified spatiotemporal database of land reserves to obtain land ownership data, and then conducting a land compliance assessment on the land ownership data to obtain land compliance assessment data, includes: Land ownership characteristics are extracted from the unified spatiotemporal database of land reserves to obtain land parcel ownership data; Land is classified based on land ownership data to obtain data on state-owned reserve land and collectively reserved land. Access the land compliance rules database, which includes land use control boundaries, constraints on historical buildings, ecological protection red line thresholds, and legal procedures for land expropriation; In accordance with land compliance rules, the data of state-owned reserve land parcels are assessed for compliance, thereby obtaining compliance assessment data of state-owned land parcels. In accordance with land compliance rules, the data of collectively reserved land parcels are used to conduct a compliance assessment of collective land parcels, thereby obtaining compliance assessment data of collective land parcels; The compliance assessment data of state-owned land parcels and collective land parcels are combined to resolve conflicts and obtain land compliance assessment data.

[0009] Optionally, the step of conducting a compliance assessment of state-owned land reserve data based on land compliance rules to obtain state-owned land compliance assessment data includes: The data of state-owned reserve land parcels are divided into planning stages according to land compliance rules, thereby obtaining data on land parcels awaiting approval and land parcels that have been approved. Based on land compliance rules, the land use matching degree of approved land parcels is calculated to obtain compliant land parcel data and conflicting land parcel data; Compliance trajectory visualization is performed on compliant land parcel data and conflict land parcel data respectively to obtain compliance trajectory map sets and conflict trajectory map sets; Based on the compliance trajectory map set, legal matching statistics are performed to obtain compliance benchmark map data for state-owned land parcels; Based on the compliance benchmark map data of state-owned land parcels, source analysis of violations is performed on the conflict trajectory map set to obtain conflict land parcel assessment data; based on the compliance benchmark map data of state-owned land parcels, compliance index is calculated on the compliance trajectory map set to obtain compliant land parcel assessment data. The compliant land assessment data and the conflict land assessment data are weighted and merged to obtain the compliant assessment data of state-owned land.

[0010] Optionally, the step of conducting a compliance assessment of the collective reserved land parcel data according to land compliance rules to obtain collective land parcel compliance assessment data includes: Based on land compliance rules, the data of collective reserved land parcels are further subdivided by land nature to obtain data on collective commercial construction land, homestead land, and unused collective land. For data on collectively owned commercial construction land, the compliance of collective land transactions is verified, including checking the land transfer filing status, the approval rate of collective members' votes, and the publicity period of the income distribution plan, thereby obtaining compliance verification data for commercial land. For homestead data, the verification of homestead eligibility rights and use rights is carried out separately, including verifying the compliance of one homestead per household, records of penalties for exceeding the area occupied, and the status of historical ownership disputes, so as to obtain homestead ownership compliance data; For data on unused collective land, a dual verification of ecological protection and arable land protection is conducted, including comparing data on permanent basic farmland protection zones, historical industrial and mining land registration data, and soil pollution risk control lists, in order to obtain environmental compliance data for unused land. Based on compliance verification data for commercial land, compliance data for homestead ownership, and compliance data for the environment of unused land, a compliance weight matrix for collective land is constructed. The weight for commercial land focuses on procedural legality, the weight for homestead emphasizes the protection of rights and interests, and the weight for unused land strengthens ecological constraints. The three types of data are weighted and fused using a weight matrix, and data from the collective member objection feedback channel is injected for dynamic correction, thereby obtaining compliance assessment data for collective land parcels.

[0011] Optionally, the step of obtaining policy change data for land reserve areas and dynamically analyzing planning constraints based on this data to obtain policy constraint data; and conducting regional land idling risk assessment based on the unified spatiotemporal database of land reserves and policy constraint data to obtain regional land idling risk data, includes: Obtain data on policy changes in land reserve areas; Policy keywords and expiration dates are extracted from data on changes in land reserve policies, thereby obtaining structured policy data. The planning adjustment range is calculated from the structured policy data to obtain the planning adjustment data; Constraint type clustering is performed on the planning adjustment data to obtain rigid constraint data and flexible constraint data; Structured policy data is dynamically analyzed based on rigid constraint data and flexible constraint data to obtain policy constraint data; Based on the unified spatiotemporal database of land reserves and policy constraint data, combined with land transaction cycle data, idle land risk modeling is performed to obtain regional land idle risk data.

[0012] Optionally, the idle land risk modeling is performed based on a unified spatiotemporal database of land reserves and policy constraint data, combined with land transaction cycle data, to obtain regional land idle risk data, including: Constructing a land transaction map based on a unified spatiotemporal database of land reserves; The constraint intensity of policy constraint data is quantified to obtain a constraint intensity index; By aligning the land transaction map with the constraint intensity index in time and space, we can obtain land-policy correlation data. A land vacancy risk prediction model is constructed based on land parcel-policy correlation data and historical land vacancy records. By using an idle risk prediction model to calculate the risk probability of land parcel-policy correlation data, regional land idle risk data can be obtained.

[0013] Optionally, the process of predicting land value fluctuations from land compliance assessment data to obtain land value prediction data; and then adjusting the land value prediction data based on regional land idling risk data using a weighted correction for the impact of idling, thereby obtaining corrected land value data, and uploading it to the land reserve management platform to perform the task of marking value anomaly blocks, includes: Land value fluctuations are predicted by analyzing land compliance assessment data, thereby obtaining land value prediction data. Based on land compliance assessment data and land compliance rules, policy change scenario simulations are conducted to obtain policy scenario simulation data. Marginal impact factors are then extracted from the policy scenario simulation data to obtain idleness impact factors. Risk-value coupling analysis is performed on regional land idling risk data and land value prediction data to obtain coupled anomaly data; A value correction model is constructed based on idle influencing factors and coupled abnormal data; The coupled abnormal data are dynamically weighted and corrected according to the value correction model to obtain corrected land value data, which is then uploaded to the land reserve management platform to perform the task of marking abnormal value blocks.

[0014] Optionally, the generation of inefficient land revitalization strategies based on regional land idling risk data and policy constraint data, thereby obtaining inefficient land revitalization strategies, is uploaded to the land reserve management platform to execute dynamic land reserve optimization tasks, including: Based on regional land idling risk data, policy constraint data is clustered to identify the causes of idling, thereby obtaining causal classification data. Development potential features are extracted from the unified spatiotemporal database of land reserves to obtain land parcel development potential data; Based on the causal classification data and the land development potential data, revitalization strategies are matched to obtain a preliminary set of revitalization strategies; The initial set of revitalization strategies is verified for economic feasibility and policy compliance to obtain strategies for revitalizing inefficient land use. These strategies are then uploaded to the land reserve management platform to execute the dynamic optimization of land reserves.

[0015] Furthermore, to achieve the above objectives, the present invention also provides a land reserve data multi-source integration and dynamic update system, the system comprising: a memory, a processor, and a land reserve data multi-source integration and dynamic update program stored in the memory and executable on the processor, the land reserve data multi-source integration and dynamic update program being configured to implement the steps of the land reserve data multi-source integration and dynamic update method as described above.

[0016] This invention provides a method for multi-source integration and dynamic updating of land reserve data. The method utilizes hierarchical coordinate fusion technology to incorporate heterogeneous data from multiple sources, such as land surveys, real estate registration, and planning permits, into a unified spatiotemporal framework. This solves the problem of traditional data silos, achieving unified spatial benchmarks, standardized element representation, and traceable temporal changes for land reserve data, providing a precise data foundation for end-to-end management. By extracting land parcel ownership characteristics from the unified spatiotemporal database and conducting compliance assessments, it can systematically screen compliance risks in land acquisition, transfer, and registration processes, incorporating legal compliance into the data assessment dimension. This reduces the potential for legal disputes related to land reserves from the source, enhancing the standardization and credibility of reserve activities. Furthermore, by capturing policy change data in real time and analyzing planning constraints, abstract policies are transformed into quantifiable spatial constraint indicators, enabling land reserve strategies to dynamically adapt to the latest regulatory requirements. This avoids decision-making deviations caused by policy lags and achieves real-time linkage between policy, data, and management. Finally, by integrating spatiotemporal data and policy constraints to construct an idle risk assessment model, it can proactively identify potential idle land parcels due to planning conflicts, unclear ownership, and other reasons, allowing for early intervention and reducing land resource waste. Compared to traditional manual inspection methods, this approach significantly improves risk identification efficiency, providing a scientific basis for revitalizing existing land. It introduces a weighted correction mechanism for idle land risk into land value prediction, changing the traditional static valuation model by incorporating dynamic factors such as land use status and policy risks into the valuation system, making the assessment results closer to market realities. Based on risk assessment and policy constraints, it generates customized revitalization plans for each land parcel, breaking through the traditional experience-driven standardized disposal model. Through data mining, it automatically matches revitalization paths, achieving intelligent management of the entire process from identification to assessment to disposal of inefficient land parcels, improving the economic benefits and space utilization efficiency of land reserves. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating an embodiment of the method for multi-source integration and dynamic updating of land reserve data according to the present invention.

[0018] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0020] Reference Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of the method for multi-source integration and dynamic updating of land reserve data according to the present invention, which presents an embodiment of the method for multi-source integration and dynamic updating of land reserve data according to the present invention.

[0021] In one embodiment, the method for multi-source integration and dynamic updating of land reserve data includes: Step S100: Obtain multi-source data of the land reserve area, and perform hierarchical data coordinate fusion on the multi-source data of the land reserve area to obtain a unified spatiotemporal database of land reserves.

[0022] Step S200: Extract land ownership characteristics from the unified spatiotemporal database of land reserves to obtain land ownership data, and conduct a land compliance assessment on the land ownership data to obtain land compliance assessment data.

[0023] The unified spatiotemporal database for land reserves can be a centralized dataset with a unified spatial coordinate system, standardized element structure, and timestamps, formed by integrating multi-source land data. It serves as a standardized data input source to support subsequent operations such as ownership analysis, policy matching, risk assessment, and value prediction. In this embodiment, the unified spatiotemporal database for land reserves can combine hierarchical data coordinate fusion technology to map raw data from different sources, formats, and spatiotemporal benchmarks to a unified reference frame, aggregating and storing it as structured geographic information data by plot unit. For example, the unified spatiotemporal database for land reserves can include, but is not limited to, one or more of the following: vector plot element sets, raster planning constraint layers, and time-series change log tables.

[0024] Hierarchical data coordinate fusion can be a method for aligning and registering heterogeneous geographic data layer by layer based on a multi-level spatial reference system. Furthermore, hierarchical data coordinate fusion can use the national geodetic coordinate system as the top-level benchmark, progressively adapting to the coordinate systems of land surveys, real estate registration parcel coordinates, and planning permit map patch coordinates. Spatial registration is achieved through control point matching, affine transformation, and grid correction, thereby enabling precise alignment of multi-source heterogeneous spatial data under a unified geographic benchmark.

[0025] Step S300: Obtain policy change data for the land reserve area, and perform dynamic analysis of planning constraints based on this data to obtain policy constraint data. Then, conduct a regional land idling risk assessment based on the unified spatiotemporal database of land reserves and the policy constraint data to obtain regional land idling risk data.

[0026] The policy constraint data can be a set of spatial rules generated through dynamic parsing of planning constraints, which can be directly referenced by a geographic information system (GIS). This allows policy changes to be directly referenced by the spatial database, enabling automatic identification and application of regulatory requirements. In this embodiment, dynamic parsing of planning constraints can combine spatial clauses in policy documents, such as changes in land use, adjustments to floor area ratio, and ecological protection red lines, and extract and map them into spatial buffers, land use coding rules, or prohibited development area layers in the GIS through natural language processing. For example, dynamic parsing of planning constraints can employ one or more of the following: land use control rule coding, development intensity threshold boundaries, and ecological red line buffer generation.

[0027] Regional land idling risk assessment can be a logical judgment system that integrates the spatial attributes of land parcels with policy constraints to determine the potential for land idling. Furthermore, regional land idling risk assessment can construct a rule engine or probability scoring model based on multi-dimensional characteristics such as land ownership integrity, consistency of planned use, compliance of development timelines, and surrounding construction progress to output the idling risk level of land parcels, thereby enabling automated identification and prioritization of potentially idle land parcels. In this embodiment, the inputs to regional land idling risk assessment include the set of land parcel elements in the unified spatiotemporal database of land reserves and policy constraint data dynamically generated through the analysis of planning constraints.

[0028] Step S400: Land value prediction is performed on the land compliance assessment data to obtain land value prediction data. Based on regional land idling risk data, the land value prediction data is weighted and corrected for the impact of idling, resulting in corrected land value data, which is then uploaded to the land reserve management platform to execute the task of marking abnormal value blocks.

[0029] The idling impact weighted correction mechanism can be a mathematical weighted algorithm that uses the idling risk level of a land parcel as a correction factor to adjust the land value prediction results. This can be used to ensure that land valuation reflects its actual usability and avoid overestimating the value of land parcels with usability obstacles. In this embodiment, the idling impact weighted correction mechanism multiplies the output of the traditional valuation model by a correction coefficient negatively correlated with the idling risk level. The higher the risk, the greater the valuation discount, resulting in a dynamically adjusted value output. For example, the idling impact weighted correction mechanism can employ a linear discount correction model, a nonlinear risk decay model, or an interval threshold weighted model.

[0030] Step S500: Generate inefficient land revitalization strategies based on regional land idling risk data and policy constraint data, thereby obtaining inefficient land revitalization strategies and uploading them to the land reserve management platform to execute the dynamic optimization task of land reserves.

[0031] Among them, the inefficient land revitalization strategy can be a customized disposal suggestion output based on risk type and policy allowance, matching historical successful cases or rule bases. It can be used to achieve end-to-end automated recommendation from problem identification to disposal solution. In this embodiment, the generation of inefficient land revitalization strategy can combine regional land idling risk data and policy constraint data to establish a strategy knowledge graph or use collaborative filtering algorithms to recommend strategy combinations based on the historical disposal behavior of similar plots, thereby achieving end-to-end automated recommendation from problem identification to disposal solution.

[0032] Taking the dynamic management of reserve land parcels in urban renewal areas as an example, the multi-source integration and dynamic updating method for land reserve data in this embodiment can be as follows: In a certain urban renewal area, the original industrial land has been adjusted to mixed commercial and residential land due to planning changes, but some parcels have not been expropriated due to ownership disputes. The system unifies the parcel boundaries through hierarchical coordinate fusion, automatically identifies new planning restrictions through dynamic analysis of policy constraints, and determines that the parcel is a high-risk idle parcel due to unclear ownership and undetermined use by the idle risk assessment model. The idle impact weighted correction mechanism lowers its valuation by 50%, and the inefficient land revitalization strategy generation module recommends a combination of expropriation + planning fine-tuning + joint investment promotion, which is then pushed to the management platform for decision-makers' reference.

[0033] This embodiment provides a method for multi-source integration and dynamic updating of land reserve data. It acquires multi-source data from land reserve areas and performs hierarchical data coordinate fusion to construct a unified spatiotemporal database. It extracts ownership characteristics and assesses compliance to form basic assessment data. It analyzes policy changes to generate calculable spatial constraints and combines them with the database to assess idle risk. It dynamically adjusts land value prediction results by using risk level as a correction factor. Finally, it generates revitalization strategies by integrating risk and policy dimensions and pushes them to the management platform. This method achieves spatial benchmark unification of multi-source heterogeneous data through hierarchical data coordinate fusion, providing a basis for subsequent analysis. A consistent data foundation; dynamic analysis of policy constraints transforms policy texts into calculable spatial rules, enabling management actions to respond in real time to regulatory changes; an idle risk assessment model integrates spatial and policy dimensions to identify potentially inefficient land parcels, breaking through the limitations of static assessment; a weighted correction mechanism for the impact of idleness embeds risk factors into the valuation system, ensuring that land value reflects its true usability; ultimately, customized revitalization strategies are generated based on both risk and policy inputs, forming a closed-loop processing flow from data fusion, risk identification, value correction to strategy output, achieving the technical effect of transforming land reserve management from passive recording to proactive adaptation, and from experience-based judgment to data-driven systemic transformation.

[0034] In one embodiment, multi-source data of the land reserve area is acquired, and hierarchical data coordinate fusion is performed on the multi-source data of the land reserve area to obtain a unified spatiotemporal database of land reserves, including: Obtain multi-source data on land reserve areas, including geographic information system data, real estate registration data, urban planning map data, and land market listing data.

[0035] The geographic information system (GIS) data can be vector or raster geographic datasets containing spatial elements such as topography, land types, and administrative divisions. These datasets provide a basic representation of land's natural attributes and spatial patterns, serving as a base map for spatial analysis. In this embodiment, the GIS data can originate from the GIS platform of the natural resources department and be stored in Shapefile, GeoJSON, or Geodatabase formats. For example, the GIS data can include, but is not limited to, one or more of the following: land type status layers, topographic elevation models, and administrative boundary vectors. Real estate registration data can be legal ownership information recording the land right holder, right type, area, use, and registration time. This information provides a legal basis for land ownership boundaries and status, and is used for compliance assessment and ownership conflict identification. In this embodiment, real estate registration data can be obtained in batches from the real estate registration system through a government data interface, containing parcel codes and right holder information in a structured table format. For example, the real estate registration data can include, but is not limited to, one or more of the following: state-owned land use right registration records, collective construction land registration records, and mortgage and seizure status markers.

[0036] Urban planning map data can be the legally mandated detailed control planning results issued by the planning department, containing rules such as land use, development intensity, and spatial control boundaries. It serves as a policy constraint input source and is used to identify planning use conflicts and development restrictions. In this embodiment, urban planning map data can be exported from the planning approval system as CAD or GIS format map layers, and then standardized and converted into semantic attribute codes. For example, urban planning map data may include, but is not limited to, one or more of the following: land use zoning layers, building density control lines, and the impact range of public service facilities. Land market listing data can be market behavior data recording public land transfers, bidding announcements, successful bidders, and transaction status. It is used to reflect land market activity and potential development intentions, and to assist in judging idle risks and value fluctuation trends. In this embodiment, land market listing data can be obtained through the government's public resource trading platform API, containing fields such as plot number, listing time, starting price, and failed auction marker. For example, land market listing data may include, but is not limited to, one or more of the following: listed but unsold plots, failed auction records, and plots that have been sold but not yet started construction.

[0037] Acquiring multi-source data for land reserve areas, including geographic information system data, real estate registration data, urban planning map data, and land market listing data, can be achieved by batch acquisition of these four heterogeneous data sources through government data interfaces or file import methods. Furthermore, acquiring multi-source data for land reserve areas can be accomplished by calling the API of the natural resources department to obtain GIS and registration data, or by crawling HTML pages of public resource trading platforms to extract listing information. This allows for comprehensive collection of data across the entire land reserve chain, covering the technical effects of natural attributes, ownership status, planning requirements, and market behavior.

[0038] Data timeliness verification and coordinate system calibration are performed on multi-source data of land reserve areas to obtain a standardized dataset.

[0039] Data timeliness verification can be a mechanism for comparing and validating the timestamps of multi-source data to ensure that the data involved in the fusion has timeliness consistency and avoid misjudgment due to outdated data. In this embodiment, data timeliness verification can extract the update time field of each data source, compare it with the latest policy release date or benchmark time, and remove data items that have not been updated. For example, data timeliness verification can include, but is not limited to, one or more of the following: update cycle compliance verification, version number comparison mechanism, and automatic marking of invalid data. Coordinate system calibration can be the process of converting geographic data under different spatial reference systems to a unified coordinate system to eliminate the offset of land parcel positions caused by coordinate system differences and ensure spatial overlay accuracy. In this embodiment, coordinate system calibration can be performed by calculating coordinate transformation parameters through common control points or overlapping areas, and spatial alignment can be achieved using a seven-parameter or three-parameter model. For example, coordinate system calibration can include, but is not limited to, one or more of the following: WGS84 to CGCS2000 transformation, local independent coordinate system projection transformation, and patch boundary resampling registration.

[0040] Verifying the timeliness and calibrating the coordinate system of multi-source data for land reserve areas can be achieved by sequentially performing timestamp validity checks and spatial coordinate system transformations. Furthermore, this verification can be accomplished by setting an update timeliness threshold to automatically filter data older than six months, or by using control point registration to achieve accurate reprojection across coordinate systems. This ensures consistency and reliability of input data in both time and space, effectively reducing fusion errors.

[0041] A standardized dataset can be a preprocessed data set that has undergone time-testing and coordinate calibration, and has a unified format, fields, encoding, and spatial benchmark. It serves as input for hierarchical partitioning and dynamic association, ensuring input consistency for subsequent operations. In this embodiment, a standardized dataset can be generated by performing field mapping, encoding standardization, null value imputation, and topological cleaning on the original data. For example, a standardized dataset can include, but is not limited to, one or more of the following: a unified attribute structure table, a standard spatial reference layer, and a cleaned feature set.

[0042] Land reserve tiers are classified based on standardized datasets to obtain primary and secondary reserve data.

[0043] The land reserve tiering can be a classification mechanism that divides reserve plots into different management levels based on land development maturity and management priority. This allows for differentiated management of reserve resources, improving resource allocation efficiency and response priority. In this embodiment, the land reserve tiering can be based on indicators such as ownership integrity, planning implementation, and market activity, setting rule thresholds for clustering or decision tree classification. For example, the land reserve tiering can include, but is not limited to, one or more of the following: primary reserve data, secondary reserve data, and reserve pools to be evaluated.

[0044] The primary reserve database can be a collection of core reserve land parcels with clear ownership, well-defined planning, and near-term development potential. These parcels are used for key monitoring and rapid transfer, supporting the implementation of key projects. In this embodiment, the primary reserve database can be selected from standardized datasets that meet the criteria of undisputed ownership, approved planning, and no history of failed auctions. For example, the primary reserve database may include, but is not limited to, one or more of the following: parcels with completed land acquisition, parcels with implemented control plans, and parcels planned for transfer in the near future. The secondary reserve database can be a collection of potential reserve land parcels with incomplete ownership, planning requiring adjustment, or unclear market expectations. These parcels are used as medium- to long-term reserves and policy pilot projects, supporting dynamic adjustments and proactive intervention. In this embodiment, the secondary reserve database can identify, from standardized datasets, parcels with pending ownership coordination, unapproved planning, or those that have experienced long-term failed auctions. For example, the secondary reserve database may include, but is not limited to, one or more of the following: parcels with incomplete land acquisition, parcels with planning requiring revision, and parcels with a sluggish market.

[0045] Land reserve tiering based on standardized datasets can be achieved by classifying standardized plots into primary and secondary reserves according to preset rules. Furthermore, land reserve tiering based on standardized datasets can be accomplished using decision tree models, with criteria such as ownership integrity, planning approval status, and market listing frequency. Alternatively, rigid rules can be set, such as classifying land as a primary reserve if it has no land seizures, approved planning, and no failed auctions. This allows for refined, tiered management of reserve resources and improves the strategic targeting of resource allocation.

[0046] Based on a unified spatiotemporal benchmark, data from primary and secondary reserve warehouses are dynamically correlated to obtain time-series correlation data for reserves.

[0047] A unified spatiotemporal benchmark can serve as a common reference framework for the spatial positioning and time stamping of all land reserve data, ensuring that cross-level and cross-source data are comparable, overlayable, and traceable in spatial location and time series. In this embodiment, the unified spatiotemporal benchmark can adopt the national CGCS2000 coordinate system and a unified timestamp system (such as UTC+8 accurate to the day) as the benchmark. For example, the unified spatiotemporal benchmark can include, but is not limited to, one or more of the following: spatial reference system, timestamp synchronization mechanism, unified projection parameters, etc.

[0048] Reserve time-series correlation data can be a dataset recording the evolution and transformation relationships of land parcels in primary and secondary reserves over time. This data reveals the dynamic evolution logic of reserve status and supports historical retrospection and trend projection. In this embodiment, reserve time-series correlation data can be bound to unique land parcel identifiers through timestamp comparison to track the complete path of a land parcel's upgrade from secondary to primary or its degradation to a state awaiting evaluation. For example, reserve time-series correlation data may include, but is not limited to, one or more of the following: upgrade path logs, downgrade trigger records, and state transition timelines.

[0049] Dynamically linking primary and secondary reserve data based on a unified spatiotemporal benchmark can be achieved by establishing a state evolution chain between the two levels of reserves using unique land parcel identifiers and timestamps. Furthermore, this dynamic linking can be accomplished by constructing state transition maps to record the time and triggering conditions for a land parcel to advance from secondary to primary status, or by using a time-series database to store land parcel state change logs. This enables dynamic tracking of reserve status, breaking through the traditional static snapshot model and supporting the technical effects of evolutionary analysis.

[0050] By integrating the ownership boundaries of the land reserve time-series data, a unified spatiotemporal database of land reserves can be obtained.

[0051] In this context, ownership boundary fusion can be an operation that intelligently aligns and resolves conflicts among three types of ownership-related spatial elements: registered boundaries, actual usage areas, and planning red lines. This is used to eliminate blind spots and legal risks caused by inconsistencies in multi-source ownership expressions. In this embodiment, ownership boundary fusion can use topological relationship analysis to identify boundary overlaps, gaps, and misalignments, and automatically correct or mark disputed areas through priority rules (such as registration validity taking precedence). For example, ownership boundary fusion may include, but is not limited to, one or more of the following: correction of overlapping registered boundaries and planning red lines, marking differences between actual usage areas and registered areas, and automatic marking of ambiguous ownership areas.

[0052] The fusion of all ownership boundaries in time-series data of reserves can be achieved through topological analysis and boundary conflict resolution of multi-source ownership spatial elements. Furthermore, this fusion can be accomplished by using a GIS topology engine to identify overlaps, gaps, and overhangs, automatically correcting them according to registration validity priority, or generating manual verification markers for boundary disputes that cannot be automatically resolved. This achieves precise and unified expression of ownership space, eliminating blind spots in reserves caused by differences in multi-source boundaries.

[0053] Taking the dynamic management of reserve land parcels in new urban areas as an example, the multi-source integration and dynamic updating method of land reserve data in this embodiment can be as follows: In a new urban area, a collective construction land parcel that was originally a secondary reserve is transferred from the secondary reserve to the primary reserve due to the completion of expropriation and approval of the control plan. The system automatically identifies the status change and migrates it from the secondary reserve to the primary reserve. The migration time and triggering conditions are recorded in the reserve time-series correlation data. At the same time, the system detects a 5-meter gap between its registered boundary and the planning red line. The ownership boundary fusion module automatically corrects the gap according to the registered data to generate a unified boundary. The updated status and precise boundary of the land parcel are then synchronized to the unified spatiotemporal database for subsequent value prediction and revitalization strategy modules to call.

[0054] This embodiment provides a method for multi-source integration and dynamic updating of land reserve data. It acquires multi-source data from land reserve areas, comprehensively collecting data on natural attributes, ownership status, planning requirements, and market behavior. Data timeliness verification and coordinate system calibration ensure the consistency and reliability of input data across time and space dimensions. Standardized datasets provide a unified format and structured input basis for hierarchical classification. Land reserve hierarchical classification enables differentiated management of core and potential reserves. A unified spatiotemporal benchmark establishes a spatiotemporal alignment framework for cross-level data. Reserve time-series correlation data reveals the dynamic evolution path of land parcel status. Integration of ownership boundaries eliminates inconsistencies in spatial expression of ownership. Ultimately, it constructs a unified spatiotemporal database for land reserves with dynamic evolution capabilities, precise ownership alignment, and clear, traceable hierarchical levels. This provides a high-fidelity, reasonable, and structured data foundation for subsequent compliance assessments, risk modeling, and intelligent decision-making.

[0055] In one embodiment, the step of extracting land ownership characteristics from the unified spatiotemporal database of land reserves to obtain land parcel ownership data, and then conducting a land compliance assessment on the land parcel ownership data to obtain land compliance assessment data, includes: Land ownership characteristics are extracted from the unified spatiotemporal database of land reserves to obtain land parcel ownership data; The unified spatiotemporal database for land reserves can be a comprehensive land management data platform integrating spatial location, ownership information, and time-series attributes, which can be used to support dynamic monitoring and compliance analysis of land resources. In this embodiment, the unified spatiotemporal database for land reserves can combine attribute fields such as land parcel ownership identifiers, owners, and registration status. Records containing state-owned / collective identifiers in the ownership fields can be filtered through SQL queries, or ownership information can be supplemented by associating real estate registration numbers with geographic information system attribute tables. This allows for the extraction of a basic set of ownership attributes for subsequent classification and compliance assessment.

[0056] Land is classified based on land ownership data to obtain data on state-owned reserve land and collectively reserved land.

[0057] The state-owned reserve land data can be a collection of land parcels extracted from the unified spatiotemporal database of land reserves, with ownership type being state-owned and used for government reserves. This data can serve as an independent evaluation unit, subject to compliance verification under state-owned land management regulations. In this embodiment, the state-owned reserve land data can be filtered based on the ownership nature identifier in the land parcel ownership field to identify land parcels registered as state-owned and included in the reserve management scope. These can include, but are not limited to, one or more of the following: expropriated but not supplied land parcels, state-owned reserve land parcels under planning adjustments, and temporary state-owned reserve land parcels.

[0058] The data on collectively reserved land parcels can be a collection of reserved land parcels extracted from the unified spatiotemporal database of land reserves, whose ownership type is collective and whose expropriation or transfer has not yet been completed. These can serve as independent evaluation units and be subject to compliance verification against regulations governing collective land management and expropriation procedures. In this embodiment, the data on collectively reserved land parcels can be filtered based on the collective ownership identifier in the ownership field, combined with the attribute tag indicating that the land use is reserved for reserves. This can identify collective land units that have not completed the nationalization process, including but not limited to one or more of the following: reserved collective land in urban villages, parcels for which expropriation compensation has not been completed, and reserved areas for collective commercial construction land.

[0059] Land classification based on land ownership data yields data on state-owned reserve land and collectively reserved land. This can be achieved by dividing land into two independent datasets—state-owned reserves and collectively reserved land—based on the legal classification standards of the ownership attribute field. Furthermore, land classification based on ownership data can be achieved through hard segmentation using enumerated values ​​of the ownership type field, or by combining land use and expropriation status fields for composite classification. This allows for the separate handling of state-owned and collectively owned land at the legal application level, avoiding rule mismatches.

[0060] Access the land compliance rules database, which includes land use control boundaries, constraints on historical buildings, ecological protection red line thresholds, and legal procedures for land expropriation.

[0061] The land compliance rule base can be a structured collection of rules that store spatial constraints and procedural requirements in laws and regulations related to land management. It can provide standardized and calculable criteria for assessing the compliance of state-owned and collectively owned land parcels. In this embodiment, the land compliance rule base can transform textual norms such as land use control, ecological protection red lines, and expropriation procedures into executable logical conditions and spatial boundary determination rules through semantic parsing of legal provisions and spatial rule mapping. These rules may include, but are not limited to, land use control boundary rule sets, rules governing the impact range of historical buildings, ecological protection red line threshold rules, and rules governing statutory procedures for land expropriation.

[0062] In accordance with land compliance rules, the data of state-owned reserve land parcels are used to conduct compliance assessments of state-owned land parcels, thereby obtaining compliance assessment data of state-owned land parcels.

[0063] Specifically, the data on state-owned reserve land parcels is matched item by item with the relevant rules on state-owned land in the land compliance rule database, and compliance status labels are output. Furthermore, the compliance assessment of state-owned reserve land parcel data based on land compliance rules can be conducted by using a rule engine to verify whether the land use exceeds limits, whether it is within a protected area, and whether the statutory expropriation procedures have been completed, or by constructing a compliance scoring matrix to assign scores to the degree of compliance of each rule and summarizing them to generate a compliance level. This allows for the generation of compliance determination results for state-owned land management regulations.

[0064] In accordance with land compliance rules, the data of collectively reserved land parcels are used to conduct a compliance assessment of collective land parcels, thereby obtaining compliance assessment data of collective land parcels.

[0065] Specifically, the data on collectively reserved land parcels is matched item by item with the relevant rules on collective land in the land compliance rule database, and a compliance status label is output. Furthermore, the compliance assessment of collectively reserved land parcel data based on land compliance rules can be performed by verifying whether there are violations of collective land use restrictions, failure to fulfill public disclosure procedures, or excessive reservations, or by combining land acquisition compensation progress data to determine whether statutory preconditions are met, thereby generating a compliance determination result for collective land management regulations.

[0066] The compliance assessment data of state-owned land parcels and collective land parcels are combined to resolve conflicts and obtain land compliance assessment data.

[0067] Conflict resolution and consolidation can be a mechanism for consistent correction of logical contradictions or spatial overlaps in the compliance assessment results of state-owned and collectively owned land parcels. This ensures that the final compliance assessment results are spatially and logically consistent and free from contradictions, making them directly applicable to administrative decision-making. In this embodiment, conflict resolution and consolidation can identify contradictions in the assessment results of two types of land parcels in scenarios such as overlapping boundaries, overlapping uses, and procedural timeliness conflicts. It uses rule-driven adjudication and result fusion based on legal priority, spatial topological relationships, and the scope of policy application. This can include, but is not limited to, spatial boundary priority adjudication mechanisms, procedural timeliness priority rules, and legal ownership validity ranking mechanisms.

[0068] Specifically, the system identifies assessment contradictions between the two types of land parcels in spatial overlap or procedural conflicts, and performs consistency correction and result fusion based on legal priority and spatial topology rules. Furthermore, conflict resolution and merging of compliance assessment data for state-owned and collectively owned land parcels can be achieved by prioritizing the planning use determination result of state-owned land parcels when their boundaries overlap, or by prioritizing the application of state-owned procedural rules based on the Land Administration Law when the same land parcel is simultaneously marked as both state-owned and collectively owned non-compliant. This eliminates self-contradictions in compliance assessments caused by mixed ownership, resulting in a unified and executable final compliance assessment result.

[0069] Taking the compliance review of reserve land parcels in urban-rural fringe areas as an example, the multi-source integration and dynamic updating method for land reserve data in this embodiment can be as follows: some parcels in a certain area were originally collective agricultural land, which were included in the scope of state-owned reserves after planning adjustments, but the expropriation process has not yet been completed. After extracting the ownership, the system classifies them as collective reserved land parcels and simultaneously identifies adjacent parcels that have been designated as state-owned reserves. The ecological protection red line rules in the compliance rule base apply to both types of parcels, but the expropriation procedure rules only apply to state-owned parcels. During the assessment, it was found that the collective parcel was incorrectly marked as having been expropriated in compliance. The system, through a conflict resolution and merging mechanism, determines that the parcel should be classified as collective non-compliance - expropriation not completed based on the priority of the statutory expropriation process, and automatically corrects the assessment results to avoid the risk of subsequent illegal land supply.

[0070] This embodiment extracts land ownership characteristics from a unified spatiotemporal database of land reserves to obtain land ownership data. Based on the land ownership data, land is classified to obtain data on state-owned reserve land and collectively reserved land. A land compliance rule base is obtained, and compliance assessments are conducted on the two types of land separately. The assessment results are then integrated through a conflict resolution and merging mechanism. This enables the structured expression of legal norms, precise separation of state-owned and collective land at the level of legal application, and systematic resolution of assessment contradictions caused by overlapping ownership and procedures. This ensures that the compliance judgment results are unique and enforceable in terms of space and legal logic, significantly improving the accuracy, consistency, and legal traceability of compliance assessments in land reserve management.

[0071] In some embodiments, the step of conducting a compliance assessment of state-owned land reserve data based on land compliance rules to obtain state-owned land compliance assessment data includes: According to land compliance rules, the data of state-owned reserve land parcels are divided into planning stages to obtain data on land parcels awaiting approval and land parcels that have already been approved.

[0072] The data on land parcels awaiting approval can be a collection of records corresponding to state-owned reserve land parcels that have not yet completed the approval process. This data can be used to support the prioritization and resource allocation of subsequent approval processes. For example, the data on land parcels awaiting approval can be categorized and stored according to approval status, application time, and regional distribution. The data on land parcels already approved can be compliance records corresponding to state-owned reserve land parcels that have completed the statutory approval process. This data can be used to support the calculation of land use matching degree and the construction of compliance trajectories. In a specific embodiment, the data on land parcels already approved may include structured fields such as approval document number, approval time, planning use code, and land area.

[0073] Based on land compliance rules, the land use matching degree of approved land parcels is calculated to obtain compliant land parcel data and conflicting land parcel data.

[0074] The compliant land parcel data can be a collection of land parcels whose land use changes or usage behaviors are completely consistent with the approved planning, and can be used to construct stable and reusable compliance behavior paths. The conflicting land parcel data can be records of land parcels with excessive land use, missing procedures, or regulatory conflicts, and can be used to identify high-frequency patterns of violations and deviations in the application of law. In an exemplary embodiment, the compliant land parcel data and the conflicting land parcel data can be compared through a rule engine to determine the coding mapping relationship between the actual land use and the approved land use, and the determination can be made in conjunction with the completeness of the legal procedures for changing land use.

[0075] Compliance trajectory visualization is performed on compliant land parcel data and conflict land parcel data respectively to obtain compliance trajectory map sets and conflict trajectory map sets.

[0076] The compliance trajectory atlas can be a graph-structured dataset constructed from the full lifecycle compliance events of compliant land parcels according to time and spatial sequences. It can present the evolution path of the compliance status of land parcels at different stages and support historical behavior pattern analysis. Furthermore, the compliance trajectory atlas can be based on the compliance records of land parcels at nodes such as planning approval, land use change, and expropriation procedures, using land parcels as nodes, compliance events as edges, and timestamps as attributes to construct a dynamically evolving graph network. For example, the compliance trajectory atlas can include one or more of the following: a planning approval node graph, a land use change compliance sequence, and an expropriation procedure execution chain. The conflict trajectory atlas can be a graph-structured dataset constructed from the violations of conflicting land parcels and their associated legal provisions according to spatiotemporal relationships. It can reveal the patterns, frequency, and root causes of legal conflicts, supporting precise attribution analysis. Furthermore, the conflict trajectory atlas can extract violation event records of conflicting land parcels, associate the triggered legal provisions, the time of occurrence, the spatial location, and the responsible party, and construct a causal graph between violations and legal rules. For example, the conflict trajectory map set may include one or more of the following: usage exceeding limit trigger map, program missing breakpoint map, boundary encroachment association map, etc. Visualizing the compliance trajectory of compliant land parcel data and conflicting land parcel data separately can be achieved by connecting the compliance events of the land parcels in chronological order into a node sequence, and storing the logical relationships and spatiotemporal attributes between events in a graph structure. Furthermore, this visualization process can be achieved by constructing a land parcel-event-regulation triplet network using the Neo4j graph database, thereby transforming static compliance records into analyzable dynamic behavioral trajectories, supporting path backtracking and pattern recognition.

[0077] By performing regulatory matching statistics based on the compliance trajectory map set, compliance benchmark map data for state-owned land parcels can be obtained.

[0078] The compliance benchmark map data for state-owned land parcels can be a three-dimensional compliance model of regulations, land parcels, and time, generated by aggregating trajectory maps of historical compliant land parcels. This model can serve as a reference for compliance judgment, used to compare and analyze the degree of deviation of conflicting land parcels. Furthermore, the compliance benchmark map data can statistically cluster all compliance paths in the compliance trajectory map set, extracting high-frequency compliance event sequences, typical approval processes, and regulatory combination patterns to form standardized compliance benchmark templates. For example, the compliance benchmark map data for state-owned land parcels may include one or more of the following: high-frequency approval process templates, standard use change path sets, and complete legal procedure chain models. Statistical matching of regulations based on the compliance trajectory map set can involve statistically analyzing recurring regulatory combinations, approval processes, and event sequences in the compliance trajectory map set, extracting high-frequency patterns, and aggregating them into benchmark templates. Further, this statistical process can use the FP-Growth algorithm to mine frequent compliance event sets, thereby enabling the construction of a reusable compliance behavior standard model as a reference benchmark for evaluating new land parcels.

[0079] Based on the compliance benchmark map data of state-owned land parcels, the source analysis of violations is carried out on the conflict trajectory map set to obtain the assessment data of conflict land parcels.

[0080] Among these, conflict land parcel assessment data can be quantitative evaluation results reflecting the severity and root cause of violations on conflict land parcels, which can be used to support differentiated handling and administrative intervention decisions. Based on the compliance benchmark map data of state-owned land parcels, a source-tracing analysis of the conflict trajectory map set can be conducted. This involves comparing the violation trajectory of the conflict land parcels with the compliance benchmark map to locate the nodes deviating from the benchmark and the violated legal provisions. Furthermore, this source-tracing analysis can use graph similarity algorithms to calculate the difference between the conflict path and the benchmark path, thereby automatically identifying the causes and key nodes of the violations and achieving precise location of the root cause of the violations.

[0081] Based on the compliance benchmark map data of state-owned land parcels, the compliance index of the compliance trajectory map set is calculated to obtain the assessment data of compliant land parcels.

[0082] The compliant land parcel assessment data can serve as a continuous score measuring the degree of matching between compliant land parcels and the benchmark map, supporting refined hierarchical management and risk warning. Calculating a compliance index on the compliance trajectory map set based on the state-owned land parcel compliance benchmark map data can measure the degree of matching between the compliant land parcel trajectory and the benchmark map, generating a continuous compliance score between 0 and 1. Furthermore, this calculation process can calculate trajectory similarity based on path edit distance, thereby transforming compliance status into a quantifiable numerical indicator to support refined hierarchical management.

[0083] The compliant land assessment data and the conflict land assessment data are weighted and merged to obtain the compliant assessment data of state-owned land.

[0084] The weighted merging mechanism can be a mathematical operation that combines the assessment scores of compliant and conflict-affected land parcels according to preset weights to form a final compliance assessment value. This enables the transformation from discrete compliance labels to continuous risk scores, supporting tiered response and priority ranking. Furthermore, the weighted merging mechanism can set asymmetric weights based on the stability of the compliance index and the severity of the conflict assessment, generating a continuous compliance score through weighted summation. For example, the weighted merging mechanism can employ one or more of the following: compliance index-driven weighting, violation severity-driven weighting, and dynamic decay weighting models. The weighted merging of compliant and conflict-affected land parcel assessment data can be achieved by weighting the compliance index and conflict severity scores according to set weights, outputting the final compliance assessment value. Further, this merging process can achieve an upgrade from binary compliance judgment to continuous risk scoring by setting a compliance index weight of 0.7 and a conflict severity weight of 0.3 for linear weighting, thus supporting the generation of differentiated handling strategies.

[0085] Taking the compliance review of state-owned reserve land in historical industrial zones as an example, the multi-source integration and dynamic updating method of land reserve data in this embodiment can be as follows: a certain land parcel has undergone three changes in land use, with the first two being compliant, and the third being converted to commercial and residential use without fulfilling the public disclosure procedure. The system identifies the first two stages as standard paths through the compliance trajectory map set, and the third deviates from the baseline map; the conflict trajectory map set locates the violation node as not being publicly disclosed; the compliance baseline map data provides a standard public disclosure process template; the compliance index is calculated to be 0.85, and the conflict score is 0.62; after weighted merging, the comprehensive compliance score is 0.73, and the system automatically marks it as a medium risk and pushes rectification suggestions to avoid subsequent administrative litigation due to procedural deficiencies.

[0086] This embodiment divides state-owned reserve land data into planning stages according to land compliance rules to obtain data on land parcels awaiting approval and those already approved; it distinguishes between compliant and conflicting land parcels through land use matching degree calculation; it constructs compliance trajectory map sets and conflict trajectory map sets through compliance trajectory visualization; it generates compliance benchmark map data for state-owned land parcels through regulatory matching statistics; it quantifies the assessment values ​​of conflicting and compliant land parcels through violation source tracing analysis and compliance index calculation; and it merges and integrates the two types of assessment results through weighted merging, thereby transforming the land compliance status from discrete events into a dynamic map expression, upgrading the assessment benchmark from rule clauses to pattern models, changing violation attribution from manual comparison to map comparison, and transforming the assessment conclusion from qualitative labels to continuous scoring. This achieves the technical effect of improving the objectivity, traceability, and interpretability of the assessment and decision-making.

[0087] In some embodiments, collective land reserve data is assessed for compliance with land compliance rules to obtain collective land compliance assessment data, including: subdividing the collective reserved land data by land use according to land compliance rules to obtain data on collective commercial construction land, homestead land, and unused collective land.

[0088] The data on collectively owned commercial construction land can be collectively owned land units selected from collectively reserved plots for non-agricultural commercial purposes. These units can be used as independent evaluation units and subject to compliance verification focused on procedural legality. In this embodiment, the data on collectively owned commercial construction land can be combined with land use attributes and ownership identifiers, along with the legal definition of commercial land under the Land Administration Law, to screen collectively owned land with market entry potential. For example, the data on collectively owned commercial construction land may include, but is not limited to, one or more of the following: reserved collective land in industrial parks, commercial service facility land, and rural tourism development land. The data on homestead land can be collectively owned land units identified from collectively reserved plots for rural villagers' residential construction. These units can be used as independent evaluation units and subject to rights status verification focused on protecting farmers' rights. In this embodiment, the data on homestead land can be based on homestead land registration attributes, household registration information, and residential distribution density to identify residential land plots that meet the definition of residential use in the Rural Homestead Management Measures. For example, homestead data may include, but is not limited to, one or more of the following: homesteads with multiple dwellings per household exceeding the standard, homesteads acquired through inheritance but not yet registered, and homesteads historically occupied but not penalized.

[0089] Unused collective land data can refer to collectively owned land that is not planned for commercial or residential use and remains undeveloped. It can be used as an independent assessment unit, subject to compliance assessments under the dual constraints of ecological protection and farmland protection. In this embodiment, unused collective land data can be obtained by cross-referencing land use status classification with planned uses, removing registered land parcels and retaining the remaining undeveloped collective land. For example, unused collective land data may include, but is not limited to, one or more of the following: ecological buffer zone collective wasteland, unused land on the edge of permanent basic farmland, and collective land before reclamation of industrial and mining wasteland.

[0090] For data on collectively owned commercial construction land, the compliance of collective land transactions is verified, including checking the land transfer filing status, the approval rate of collective members' votes, and the publicity period of the income distribution plan, thereby obtaining compliance verification data for commercial land. Specifically, verifying the compliance of collective land transactions for commercial construction can involve checking whether the land transfer has been registered, approved by a members' meeting, and whether the profit distribution plan has been publicly disclosed. Furthermore, verifying the compliance of collective land transactions for commercial construction can be achieved by cross-referencing the natural resources department's registration system with village meeting minutes, or by verifying the integrity and timestamps of electronic signatures on votes through a blockchain-based evidence storage platform. This ensures that the process of commercial land entering the market complies with legal procedures.

[0091] For homestead data, the verification of homestead eligibility rights and use rights is carried out separately, including verifying the compliance of one homestead per household, records of penalties for exceeding the area occupied, and the status of historical ownership disputes, so as to obtain homestead ownership compliance data; Specifically, regarding homestead data, implementing the separation verification of homestead eligibility rights and usage rights can involve verifying whether farmers meet the "one household, one homestead" standard, whether there are any unpunished records of over-occupied areas, and whether there are any unresolved historical ownership disputes. Furthermore, regarding homestead data, this separation verification can be achieved by linking the public security household registration system with the homestead registration system to match household and homestead consistency, or by reviewing historical dispute files from courts or mediation committees to determine the status of ownership disputes, thereby identifying potential risks of unclear homestead rights or damage to rights.

[0092] For data on unused collective land, a dual verification of ecological protection and arable land protection is conducted, including comparing data on permanent basic farmland protection zones, historical industrial and mining land registration data, and soil pollution risk control lists, in order to obtain environmental compliance data for unused land. Specifically, for data on unused collective land, a dual verification of ecological and arable land protection can be conducted. This can involve comparing whether the plot is located within a permanent basic farmland protection zone, whether it falls within the scope of historical industrial and mining land registration, and whether it is on the soil pollution risk control list. Furthermore, for data on unused collective land, this dual verification of ecological and arable land protection can be spatially overlaid by superimposing the arable land protection red line layer from the Third National Land Survey with the list of polluted plots from the Ministry of Ecology and Environment, or by using soil environmental monitoring data to determine whether the plot is on the edge of a control zone. This can help identify legal compliance risks related to ecological and arable land protection for unused land.

[0093] Based on compliance verification data for commercial land, compliance data for homestead ownership, and compliance data for the environment of unused land, a compliance weight matrix for collective land is constructed. The weight for commercial land focuses on procedural legality, the weight for homestead emphasizes the protection of rights and interests, and the weight for unused land strengthens ecological constraints. Specifically, based on compliance verification data for commercial land, compliance data for homestead ownership, and environmental compliance data for unused land, a collective land compliance weight matrix is ​​constructed. This can involve setting weight coefficients for three assessment dimensions—procedural, rights-based, and ecological—for the three types of land parcels, forming an independent weighted assessment structure. In this embodiment, the collective land compliance weight matrix can be determined using an expert scoring method, with 60% weight for procedural aspects of commercial land, 55% for rights-based homesteads, and 70% for ecological aspects of unused land. Alternatively, it can be achieved through backtesting analysis of historical compliance dispute cases to backfit the optimal weight combination, thereby enabling differentiated weight allocation for the three types of collective land compliance assessments and avoiding homogenization of rule application.

[0094] The three types of data are weighted and fused using a weight matrix, and data from the collective member objection feedback channel is injected for dynamic correction, thereby obtaining compliance assessment data for collective land parcels.

[0095] Specifically, the three types of data are weighted and fused using a weight matrix, and then dynamically corrected by injecting data from the collective member objection feedback channel. This can be achieved by weighting and summing the compliance scores of the three types of land parcels according to the weight matrix, and then adding a correction factor for negative opinions from the objection feedback data to adjust the final evaluation value. In this embodiment, the collective member objection feedback channel data can be records of objections and feedback data regarding land disposal submitted by members of the collective organization through legal channels. It can also be collected through village-level public disclosure platforms or village representative meeting minutes, gathering objections and questions from collective members regarding land entering the market, expropriation, and allocation. Furthermore, the weighted fusion of the three types of data using a weight matrix and the dynamic correction by injecting data from the collective member objection feedback channel can be achieved by multiplying the number of objections by a preset penalty coefficient and deducting it from the total score, or by introducing an objection sentiment intensity score and generating a dynamic decay factor through text sentiment analysis. This allows for the linkage between administrative compliance assessment and grassroots social feedback, enabling dynamic calibration of the assessment results and governance response.

[0096] Taking the compliance assessment of collectively owned commercial construction land in suburban areas before it enters the market as an example, the multi-source integration and dynamic updating method of land reserve data in this embodiment can be as follows: A village plans to use a piece of collective land for the construction of logistics warehouses. The system identifies it as commercial construction land. The compliance verification shows that its voting pass rate meets the standard, but the income distribution plan has not been publicized, and the score is 0.72. At the same time, many villagers submit objections through online channels, claiming that the income is not distributed equally according to the population. The procedural legality weight in the weight matrix is ​​0.6, and the objection feedback factor triggers a penalty correction of 0.15. The final assessment value drops to 0.57, and the system automatically triggers the process of supplementary publicity and re-voting.

[0097] This embodiment subdivides the data of collectively reserved land parcels according to land compliance rules, obtaining data on collectively owned commercial construction land, homestead land, and unused collectively owned land. For each of these three types of data, it performs compliance verification for market entry and transaction, verification of the separation of eligibility rights and usage rights, and dual verification of ecological protection and arable land protection. It constructs a collectively owned land compliance weight matrix based on procedural legality, rights protection, and ecological constraints. By weighted fusion of the three types of compliance data and incorporating feedback data from collective members' objections, the system is dynamically corrected. This allows for typified weight allocation and linkage with social feedback in the evaluation process. Through structured modeling, policy objectives are transformed into calculable evaluation dimensions, integrating administrative norms with real grassroots demands. This constructs a four-dimensional linkage evaluation system encompassing policy rules, land parcel types, weight allocation, and social feedback. This effectively bridges the implementation gap between legal norms and grassroots realities, fundamentally resolving compliance misjudgments and governance conflicts caused by ambiguous property rights, opaque procedures, and the absence of public participation. This significantly enhances the legal basis and social acceptance of collectively owned land reserves.

[0098] In some embodiments, policy change data for land reserve areas is acquired, and planning constraints are dynamically analyzed based on this data to obtain policy constraint data; regional land idling risk assessment is conducted based on the unified spatiotemporal database of land reserves and the policy constraint data to obtain regional land idling risk data, including: Obtain data on policy changes in land reserve areas; Policy keywords and expiration dates are extracted from data on changes in land reserve policies, thereby obtaining structured policy data. The planning adjustment range is calculated from the structured policy data to obtain the planning adjustment data; Constraint type clustering is performed on the planning adjustment data to obtain rigid constraint data and flexible constraint data; Structured policy data is dynamically analyzed based on rigid constraint data and flexible constraint data to obtain policy constraint data; Based on the unified spatiotemporal database of land reserves and policy constraint data, combined with land transaction cycle data, idle land risk modeling is performed to obtain regional land idle risk data.

[0099] The extraction of policy keywords and marking of expiration dates for land reserve area policy change data can be achieved by using natural language processing technology to identify key semantic units in the policy text and mark their time validity range. Further, this extraction can be done by matching keywords such as "must not," "must," and "limited time" based on rule templates and associating them with date expressions, or by using a BERT model to semantically label policy paragraphs to extract the action subject, constraint object, and time boundary. This transforms ambiguous policy text into structured data units with time dimension and semantic clarity. The structured policy data, formed after keyword extraction and expiration date marking, can be a collection of policy information with timestamps, semantic tags, and spatial association attributes. This data can provide calculable and traceable policy input units for subsequent planning adjustment calculations and constraint type identification. In this embodiment, the structured policy data can be combined with named entity recognition and time expression parsing to extract structured fields such as "use adjusted to residential land" and "valid for three years from the date of issuance," and bind them to effective and expiration time windows. For example, structured policy data may include, but is not limited to, one or more of the following: mandatory use change records, development timeline constraint entries, and excerpts of compensation mechanism descriptions.

[0100] Calculating the magnitude of planning adjustments to structured policy data can quantify the impact on land development intensity or land use scope based on numerical or range information in the policy description. Furthermore, this calculation can be achieved by assigning an adjustment difference of 1.0 to statements such as "increase in floor area ratio from 1.5 to 2.5," or by extracting an upper limit threshold as the adjustment boundary for statements such as "compatible commercial proportion not exceeding 30%." This allows for a quantitative expression of policy impact, providing a comparable quantitative basis for subsequent constraint classification.

[0101] Clustering of constraint types in planning adjustment data can be performed based on the adjustment range, semantic strength, and legal effect characteristics, classifying policy impacts into rigid and flexible categories. Further, this clustering can be achieved using the K-means clustering algorithm with semantic strength scores and adjustment ranges as feature vectors for unsupervised classification, or by setting thresholds based on expert rules: if absolute terms such as "prohibited" or "must" are present, the data is classified as rigid; otherwise, it is classified as flexible. This allows for a binary distinction of policy impacts, supporting the establishment of differentiated response mechanisms. Rigid constraint data can be non-negotiable and unavoidable spatial use restrictions caused by policy changes, which can directly trigger judgments of land compliance failure and are high-weight constraint factors in idle risk assessment. In a specific embodiment, rigid constraint data can be combined with semantic classification to identify absolute expressions such as "prohibited development," "must be retained," and "cannot be changed," mapping them to impenetrable spatial restricted areas or mandatory use codes in GIS. For example, rigid constraint data can include, but is not limited to, one or more of the following: overlapping areas of ecological protection red lines, reserved land for public facilities, and historical landscape control areas. The flexible constraint data can be adjustable and conditionally adaptable space use guidance restrictions triggered by policy changes. These restrictions can be used to assess the development potential of land parcels under policy adaptation and to mitigate the risk of vacancy. In an exemplary embodiment, the flexible constraint data can be identified from structured policy data containing statements with floating ranges or conditional triggers, such as "the plot ratio can be increased to 2.5" or "approval is required after traffic assessment," and transformed into quantifiable adjustment ranges or conditional trigger rules. For example, the flexible constraint data may include, but is not limited to, one or more of the following: plot ratio floating ranges, gradient of supporting construction requirements, and phased development permit windows.

[0102] Dynamically analyzing structured policy data based on both rigid and flexible constraint data can involve combining these two types of constraints to apply spatial rules and prioritize them. Furthermore, this dynamic analysis can be achieved by first applying rigid constraints to filter out undevelopable land parcels, then applying flexible constraints to score the development potential of the remaining parcels, or by constructing a constraint priority tree with rigid constraints as the root node and flexible constraints as leaf nodes, enabling layer-by-layer conditional judgment. This improves the logic and efficiency of policy analysis and avoids misinterpreting flexible constraints as mandatory restrictions.

[0103] Based on a unified spatiotemporal database of land reserves and policy constraint data, idle risk modeling can be conducted by integrating spatial attributes, policy constraint types, and land transaction activity data to construct an idle probability prediction model. Furthermore, idle risk modeling based on the unified spatiotemporal database of land reserves and policy constraint data, combined with land transaction cycle data, can employ a logistic regression model with inputs including the number of rigid constraints, the leniency of flexible constraints, the duration of the most recent transaction, and the frequency of surrounding transactions. Alternatively, a Bayesian network can be constructed with policy changes as the dependent variable, transaction stagnation and ownership status as intermediate variables, and idleness as the outcome variable. This establishes a causal relationship between policy changes and market responses, enabling proactive identification of potential idleness. Idle risk modeling can be a multi-dimensional risk probability calculation framework that integrates land spatial attributes, policy constraint types, and market transaction cycle characteristics. This framework can be used to quantify the causal effects of policy changes on land use delays, improving the accuracy and foresight of idleness identification. For example, idle risk modeling can include, but is not limited to, one or more of the following: policy shock response models, transaction stagnation delay models, and supply-demand mismatch models.

[0104] Taking the land reserve response following the adjustment of industrial land use policies in a new district as an example, the multi-source integration and dynamic updating method for land reserve data in this embodiment can be as follows: A new district issues a policy to adjust the original industrial land to strategic emerging industry land, allowing the plot ratio to increase from 1.2 to 2.0, but requiring that the proportion of supporting R&D facilities be no less than 40%. The system extracts this policy as structured data, calculates the plot ratio adjustment range as 0.8, and identifies through clustering that the change of industrial use is a rigid constraint, while the R&D facility ratio is a flexible constraint. Combining the fact that this plot has no transaction records in the past two years and the average transaction cycle of similar plots in the surrounding area is 18 months, the idle risk model outputs a high-risk probability. This result triggers the value correction and strategy generation module, recommending a combination of time-limited investment promotion and pre-approval for R&D facilities.

[0105] This embodiment acquires policy change data for land reserve areas, extracts policy keywords, and marks expiration dates to form structured policy data. It quantifies policy impact through planning adjustment calculations, distinguishes between rigid and flexible constraints through constraint type clustering, and generates policy constraint data through hierarchical dynamic analysis of the structured policy data by combining these two types of constraints. Furthermore, it constructs an idle risk model by integrating a unified spatiotemporal database of land reserves, policy constraint data, and land transaction cycle data. This transforms policy text into structured input with temporal dimension and semantic clarity, enabling hierarchical quantification and logical response to policy impact. It establishes a causal chain of policy changes, spatial constraints, market response, and idle probability, achieving early warning of idle behavior caused by delayed policy adaptation or slow market reaction. This promotes the technological advancement of land reserve management from passive policy execution to proactive risk prediction.

[0106] In some embodiments, the step of modeling idle land risk based on a unified spatiotemporal database of land reserves and policy constraint data, combined with land transaction cycle data, to obtain regional land idle risk data includes: Constructing a land transaction map based on a unified spatiotemporal database of land reserves; The land transaction map can be a spatiotemporal network structure built with land parcels as nodes and transaction behaviors as edges, reflecting the linkage and circulation inertia of transactions between land parcels. It can be understood that the land transaction map can extract transaction records of land parcels from the unified spatiotemporal database of land reserves, construct transaction relationships between nodes according to time series, and superimpose spatial proximity and time interval weights to form a weighted graph. In this embodiment, the land transaction map can include, but is not limited to, one or more of the following: direct transaction association network, spatial proximity transmission network, and time-lag response network.

[0107] The constraint intensity of policy constraint data is quantified to obtain the constraint intensity index.

[0108] The constraint strength index can be a quantitative indicator that continuously and numerically expresses the degree of impact of policy constraints. It can be understood that the constraint strength index is a single continuous value ranging from 0 to 1, generated based on the type and scope of rigid and flexible constraints through semantic weighting and spatial coverage density calculation. For example, the constraint strength index can be obtained by using the entropy weighting method to weight and sum the semantic strength and spatial coverage of rigid and flexible constraints, or by using an expert scoring-standardization method to map the constraint type to a continuous value in the range of 0.1 to 1.0.

[0109] By aligning the land transaction map with the constraint intensity index in time and space, we can obtain land-policy correlation data.

[0110] Specifically, the land parcel-policy association data is a coupled dataset formed by spatially and temporally matching land parcel transaction graph nodes with their corresponding policy constraint intensity indices within a unified spatiotemporal framework. It can be understood that the land parcel-policy association data, through spatial overlay and temporal window alignment, binds the transaction graph features of each land parcel to its policy constraint intensity index, forming a timestamped joint record of land parcels and policies. In a specific embodiment, the land parcel-policy association data may include, but is not limited to, one or more of the following: high-constraint-low-transaction land parcel sets, low-constraint-high-stagnation land parcel sets, and time-delay response land parcel sets.

[0111] A land vacancy risk prediction model is constructed based on land parcel-policy correlation data and historical land vacancy records.

[0112] The idle land risk prediction model is a machine learning prediction system trained on land parcel-policy correlation data and historical idle land records. Understandably, this model employs supervised learning methods, using historically idle land parcels as positive samples and non-idle land parcels as negative samples. Input features include transaction graph centrality, constraint strength index, mean transaction interval, and activity level of surrounding land parcels. Furthermore, the idle land risk prediction model can use graph convolutional neural networks to model the location and constraint influence of land parcels within the network, or construct an XGBoost classification model, with input features including graph centrality, constraint index, and transaction interval.

[0113] By using an idle risk prediction model to calculate the risk probability of land parcel-policy correlation data, regional land idle risk data can be obtained.

[0114] The idle land risk prediction model is a machine learning prediction system trained on land parcel-policy correlation data and historical idle land records. Understandably, this model employs supervised learning methods, using historically idle land parcels as positive samples and non-idle land parcels as negative samples. Input features include transaction graph centrality, constraint strength index, mean transaction interval, and activity level of surrounding land parcels. Furthermore, the idle land risk prediction model can use graph convolutional neural networks to model the location and constraint influence of land parcels within the network, or construct an XGBoost classification model, with input features including graph centrality, constraint index, and transaction interval.

[0115] By using an idle risk prediction model to calculate the risk probability of land parcel-policy correlation data, regional land idle risk data can be obtained.

[0116] Understandably, this operation inputs the land parcel-policy correlation data of all current land parcels into a trained prediction model and outputs idle probability values ​​in batches. Furthermore, this operation can employ a sliding window mechanism to perform rolling predictions on the latest data, or combine it with an uncertainty estimation module to output probability confidence intervals.

[0117] Taking the risk warning of idle land after policy adjustments in urban fringe areas as an example, the multi-source integration and dynamic updating method of land reserve data in this embodiment can be as follows: A policy is issued in an urban fringe area to adjust the original warehousing land into a logistics hub, but the land area is required to be no less than 10 hectares and must be equipped with a smart sorting center. The system extracts the constraint strength index generated by this policy, which is 0.82. Combined with the historical transaction map, it is found that the transaction frequency of land parcels in this area has decreased by 40%, and the average transaction interval of surrounding land parcels has been extended to 21 months. The system performs spatiotemporal alignment of the map centrality, constraint index and transaction lag of this land parcel, and after inputting it into the idle risk prediction model, the output risk probability is 0.79, triggering a high-risk marker. Based on this, the management platform initiates targeted investment promotion and flexible release of floor area ratio plans to avoid long-term idleness caused by policy adaptation delays.

[0118] This embodiment reveals the networked characteristics of land market behavior by constructing a land transaction map based on a unified spatiotemporal database of land reserves. It achieves continuous numerical expression of policy impact by quantifying the constraint intensity of policy constraint data. By spatiotemporally aligning the land transaction map with the constraint intensity index, it establishes a coupling relationship between market behavior and policy intervention. An end-to-end probability prediction model is built based on land-policy correlation data and historical idle records. This model calculates the risk probability of land-policy correlation data and outputs a regional-scale risk distribution. This approach, by constructing a land transaction map, reveals the networked characteristics of land market behavior, transforming fragmented transactions into... The system features analyzable liquidity patterns; the constraint strength index achieves continuous quantitative expression of policy impact for the first time, enabling policy variables to be used as model inputs; the spatiotemporal alignment mechanism establishes a direct coupling relationship between land market behavior and policy intervention, breaking through the traditional fragmented analysis paradigm; the idle risk prediction model integrates graph structure, policy intensity, and historical idle patterns to construct a machine learning-driven prediction link for policy shocks, transaction suppression, and idle probability, enabling forward-looking and probabilistic identification of potential idleness, and promoting land reserve management from experience-dependent passive response to proactive intervention based on data causal inference, providing precise and calculable risk inputs for value correction and intelligent strategy generation.

[0119] In one embodiment, the process of predicting land value fluctuations from land compliance assessment data to obtain land value prediction data, and then applying a weighted correction based on regional land idling risk data to the land value prediction data to obtain corrected land value data, which is then uploaded to the land reserve management platform to perform the task of marking value anomaly blocks, includes: predicting land value fluctuations from land compliance assessment data to obtain land value prediction data; simulating policy change scenarios based on land compliance assessment data and land compliance rules to obtain policy scenario simulation data, and extracting marginal impact factors based on the policy scenario simulation data to obtain idling impact factors; performing risk-value coupling analysis on regional land idling risk data and land value prediction data to obtain coupled anomaly data; constructing a value correction model based on the idling impact factors and coupled anomaly data; and dynamically weighting and correcting the coupled anomaly data based on the value correction model to obtain corrected land value data, which is then uploaded to the land reserve management platform to perform the task of marking value anomaly blocks.

[0120] The simulation of policy change scenarios based on land compliance assessment data and land compliance rules can be achieved by using the compliance assessment results as a baseline, inputting policy rule variables, and generating multiple sets of simulated outputs of land use constraint states under future policy environments. Furthermore, this operation can be performed by using the Monte Carlo method to generate a sequence of policy parameter perturbations, simulating the evolution of compliance states one by one, or by constructing a rule state machine to trigger preset policy intervention paths based on compliance shortcomings. This allows for a systematic and repeatable extrapolation of the impact of policy changes on land use constraints. In this embodiment, the policy scenario simulation data can be a virtual dataset generated by simulating changes in policy clauses, describing the state of land use constraints under different future policy environments. This dataset can be used to quantify the potential impact of policy changes on land use availability, supporting causal inferences about marginal effects. For example, the policy scenario simulation data can include one or more scenarios such as strengthened land use restrictions, extended development timelines, and ownership freeze simulations.

[0121] Extracting marginal impact factors from policy scenario simulation data can involve comparing changes in land idling risk before and after policy simulation to calculate the incremental contribution of policy variables to the idling probability. Further, this operation can be achieved using the difference method: the difference in idling scores before and after the policy is used as the marginal factor; or using the regression residual method: after controlling for other variables, the coefficient of the policy variable is used as the marginal impact, thus transforming the abstract policy impact into quantifiable and reusable numerical factors. In this embodiment, the idling impact factor can be a quantitative indicator extracted from policy scenario simulation data, representing the marginal impact of policy changes on the land idling probability. It can be used to transform the indirect effects of policy disturbances into numerical factors that can be embedded in valuation models, realizing the calculability of the policy transmission path. For example, the idling impact factor may include one or more of the following: planning adjustment sensitivity coefficient, ownership freeze amplification factor, and approval delay cumulative weight.

[0122] The regional land idling risk data can be an assessment result describing the current likelihood of land parcels being idle, based on a unified spatiotemporal database and policy constraint data. This data can serve as a core input for risk-value coupling analysis, characterizing the uncertainty of land parcel usage under real-world constraints. In this embodiment, the regional land idling risk data can be combined with one or more of the following: a list of high-risk land parcels, a distribution map of medium-risk land parcels, and stable areas of low-risk land parcels, to form a spatial risk distribution map.

[0123] Risk-value coupling analysis of regional land idling risk data and land value prediction data can simultaneously analyze the distribution relationship between idling risk and value prediction at the spatial unit level, identifying systematic deviation patterns. Further, this operation can be achieved by constructing two-dimensional scatter plot clustering to identify anomalous clusters such as high-risk overvaluation, or by using spatial autocorrelation analysis (such as Moran's I) to detect spatial clustering of risk-value mismatches, thereby revealing structural biases in value assessment that are not captured by traditional models. In this embodiment, coupled anomalous data can be a set of data points identified in both the risk and value dimensions that exhibit unexpected deviation patterns. This data can be used to reveal systematic mismatches between value assessment and risk status, serving as training and calibration samples for the value correction model. For example, coupled anomalous data can include one or more of the following: high-risk overvaluation anomalous groups, low-risk undervaluation anomalous groups, and concentrated areas of risk-value divergence.

[0124] Constructing a value correction model based on idle influencing factors and coupled anomaly data can involve using idle influencing factors as input variables and coupled anomaly data as output targets, training a mapping function to correct valuation biases. Further, this operation can be achieved by employing a linear regression model, using influencing factors as weights to fit the valuation offset of anomaly groups, or by using random forest regression, fusing multiple influencing factors and spatial feature prediction correction coefficients, thereby establishing a calculable response chain for policy disturbances, risk changes, and value deviations. In this embodiment, the value correction model can be a mathematical function system that integrates idle influencing factors and coupled anomaly data to dynamically adjust land valuation results. It can be used to achieve nonlinear, contextualized correction of land value, enabling valuation to respond to policy disturbances and changes in risk structure. For example, the value correction model can employ one or more of the following: linear weighted regression model, gradient boosting tree corrector, neural network coupling regulator, etc.

[0125] Dynamically weighting and correcting coupled anomaly data according to the value correction model can be achieved by inputting the coupled anomaly data into the value correction model and outputting a corrected valuation after policy-risk dual weighting. Furthermore, this operation can be achieved by multiplying the correction coefficients output by the model for each anomaly plot, or by using segmented correction: applying different correction function curves according to the risk level, thereby enabling land value assessment to achieve real-time adaptive response to policy transmission and risk status. In this embodiment, the corrected land value data is the final valuation result generated after the above dynamic weighting and is used to upload to the land reserve management platform. Its generation process fully embodies the dual correction logic of policy transmission and risk coupling.

[0126] Taking the dynamic calibration of land reserve valuation under urban renewal policy adjustments as an example, the multi-source integration and dynamic updating method for land reserve data in this embodiment can be as follows: In a certain area, due to a newly introduced ecological protection policy, a former industrial reserve plot is included in a restricted development area. Using land compliance assessment data as a baseline, the system simulates a policy tightening scenario, identifying a 15% increase in the risk of ownership freezing. Through marginal impact factor extraction, the marginal contribution of the policy to the probability of idling is calculated to be 0.28. In the risk-value coupling analysis, the plot exhibits a high-risk, high-valuation anomaly pattern. Based on this, the value correction model outputs a correction coefficient of 0.65, lowering the original valuation by 35%, triggering a value anomaly marker, and providing accurate basis for subsequent planning adjustments and investment attraction strategies.

[0127] This embodiment uses land compliance assessment data to drive policy scenario simulation, extracts the marginal impact factors of policy changes on the probability of idling, and combines regional land idling risk data with land value prediction data for coupled analysis to identify systematic valuation biases. Based on these biases and influencing factors, a dynamic correction model is constructed to perform dual weighted calibration on the valuation results. This transforms land value assessment from a single financial indicator into a multi-dimensional dynamic response system that integrates legal constraints, policy changes, and utilization status. It enables real-time decoupling and calibration of the policy-risk transmission chain in value assessment, supporting the causal explanatory power and forward-looking judgment capability of value anomaly markers.

[0128] In one embodiment, an inefficient land revitalization strategy is generated based on regional land idling risk data and policy constraint data, thereby obtaining the inefficient land revitalization strategy, which is then uploaded to the land reserve management platform to execute the dynamic optimization task of the land reserve, including: Based on regional land idling risk data, policy constraint data is clustered to identify the causes of idling, thereby obtaining causal classification data. Development potential features are extracted from the unified spatiotemporal database of land reserves to obtain land parcel development potential data; Based on the causal classification data and the land development potential data, revitalization strategies are matched to obtain a preliminary set of revitalization strategies; The initial set of revitalization strategies is verified for economic feasibility and policy compliance to obtain strategies for revitalizing inefficient land use. These strategies are then uploaded to the land reserve management platform to execute the dynamic optimization of land reserves.

[0129] Among them, idle land cause clustering can be an unsupervised clustering analysis method based on multidimensional idle land risk signals to group policy and ownership conflict patterns. It can be used to reduce scattered idle land phenomena to identifiable cause categories and support targeted matching of strategy libraries. In this embodiment, idle land cause clustering can input feature vectors such as planning conflict type, ownership missing status, and approval delay cycle from regional land idle land risk data into the clustering algorithm to automatically identify high-frequency combination patterns and classify cause categories. For example, idle land cause clustering can include, but is not limited to, one or more of the following: planning use conflict type, ownership chain break type, and approval process blockage type. Idle land cause clustering based on regional land idle land risk data and policy constraint data can be performed by combining the policy conflict dimension and ownership status dimension in the idle land risk data into feature vectors and applying an unsupervised clustering algorithm to group patterns. Furthermore, based on regional land idling risk data, clustering of idling causes based on policy constraint data can be achieved by using the K-means algorithm to perform vector clustering of planning conflict types and ownership gaps, thereby enabling automated classification of idling causes and forming reusable structural attribution labels.

[0130] Revitalization strategy matching can be an intelligent matching mechanism that selects suitable solutions from a strategy knowledge base based on causal classification and development potential characteristics. This can be used to transform strategy generation from experience-based recommendations to data-driven, precise adaptation. In this embodiment, revitalization strategy matching can construct a causal-potential two-dimensional matching matrix, and output a candidate strategy set by matching historical successful cases or preset strategy templates through similarity calculation. For example, revitalization strategy matching can employ a matching degree weighted ranking algorithm, strategy rule engine triggering, collaborative filtering recommendation models, etc. Revitalization strategy matching based on causal classification data and land parcel development potential data can be achieved by using causal categories and development potential scores as input conditions to retrieve historical matching records or preset rules from the strategy knowledge base. Furthermore, revitalization strategy matching based on causal classification data and land parcel development potential data can be achieved by constructing a two-dimensional matching matrix and calculating the fit between the strategy template and the current land parcel using cosine similarity, thereby upgrading strategy recommendation from single-rule matching to multi-dimensional feature collaborative recommendation.

[0131] Economic feasibility verification and policy compliance verification can be an automated verification process that performs dual verification of the financial operability and legal compatibility of preliminary revitalization strategies. This can be used to filter out strategies that lack implementation basis or violate regulations, ensuring the feasibility of the output plan. In this embodiment, economic feasibility verification and policy compliance verification can simulate and calculate economic parameters such as the rate of return on investment, funding sources, and development cycle involved in the strategy, while comparing them with mandatory clauses in current regulations regarding land use changes, transaction restrictions, and tax policies. For example, economic feasibility verification and policy compliance verification may include, but is not limited to, net present value simulation of cash flow, compliance checks for land use changes, and verification of approval authority matching. Performing economic feasibility verification and policy compliance verification on the preliminary revitalization strategy set can be achieved by automatically comparing and simulating the economic parameters and regulatory clauses involved in the strategy. Furthermore, economic feasibility verification and policy compliance verification on the preliminary revitalization strategy set can be achieved by calling financial models to calculate the expected IRR and payback period of the strategy, and setting thresholds to filter infeasible items. This allows for dual constraint filtering before strategy output, ensuring that the recommended plan has both legality and economic operability.

[0132] Taking the intelligent revitalization of old industrial zone transformation sites as an example, the multi-source integration and dynamic updating method of land reserve data in this embodiment can be as follows: a plot of land in an old industrial zone has been redeveloped for a long time due to planning adjustments to commercial and residential land. The system identifies the cause of its idleness as an approval process blockage through clustering of idle causes, and simultaneously extracts its development potential characteristics from a unified spatiotemporal database, showing that it is near a subway station and its plot ratio can be increased; the revitalization strategy matching module calls historical similar cases and recommends a combination of government acquisition + targeted transfer + supporting facilities; the economic feasibility verification module simulates and finds that the project's IRR is lower than the threshold, and the policy compliance verification module finds that land transfer fees need to be paid; the system automatically corrects the strategy to introduce social capital for joint development + phased land supply, and finally outputs a legal and feasible customized solution.

[0133] This embodiment structures regional land idling risk data into a categorizable driving model through idleness cause clustering. It breaks through the traditional single-dimensional judgment of ownership and use by extracting development potential features. It establishes an intelligent recommendation mechanism with two dimensions of cause and potential through revitalization strategy matching. It constructs a dual-constraint filtering closed loop through economic feasibility verification and policy compliance verification. This can upgrade the disposal of inefficient land use from an experience-driven trial-and-error model to a complete automated decision-making chain based on data attribution, potential assessment, strategy matching and compliance verification. It achieves the technical effect of a systematic leap in strategy generation from extensive recommendation to accurate, compliant and executable.

[0134] Furthermore, to achieve the above objectives, the present invention also provides a land reserve data multi-source integration and dynamic update system, the system comprising: a memory, a processor, and a land reserve data multi-source integration and dynamic update program stored in the memory and executable on the processor, the land reserve data multi-source integration and dynamic update program being configured to implement the steps of the land reserve data multi-source integration and dynamic update method as described above.

[0135] In addition, to achieve the above objectives, the present invention also provides a medium storing a multi-source integration and dynamic update program for land reserve data, wherein when the multi-source integration and dynamic update program for land reserve data is executed by a processor, it implements the steps of the multi-source integration and dynamic update method for land reserve data as described above.

[0136] Other embodiments or specific implementations of the land reserve data multi-source integration and dynamic update system described in this invention can be referred to the above-mentioned method embodiments, and will not be repeated here.

[0137] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for multi-source integration and dynamic updating of land reserve data, characterized in that, The method includes: Acquire multi-source data of land reserve areas and perform hierarchical data coordinate fusion on the multi-source data of land reserve areas to obtain a unified spatiotemporal database of land reserves; Land ownership characteristics are extracted from the unified spatiotemporal database of land reserves to obtain land ownership data, and land compliance assessment is conducted on the land ownership data to obtain land compliance assessment data. Obtain policy change data for land reserve areas, and perform dynamic analysis of planning constraints based on this data to obtain policy constraint data; conduct regional land idling risk assessment based on the unified spatiotemporal database of land reserves and policy constraint data to obtain regional land idling risk data. Land value fluctuations are predicted based on land compliance assessment data to obtain land value prediction data; the land value prediction data is then adjusted for the impact of idle land based on regional land idle risk data to obtain corrected land value data, which is then uploaded to the land reserve management platform to perform the task of marking abnormal value blocks. Based on regional land idling risk data and policy constraint data, strategies for revitalizing inefficient land use are generated, and these strategies are uploaded to the land reserve management platform to execute dynamic optimization tasks for land reserves.

2. The method for multi-source integration and dynamic updating of land reserve data as described in claim 1, characterized in that, The process of acquiring multi-source data on land reserve areas and performing hierarchical data coordinate fusion on the multi-source data to obtain a unified spatiotemporal database for land reserves includes: Obtain multi-source data on land reserve areas, including geographic information system data, real estate registration data, urban planning map data, and land market listing data; Data timeliness verification and coordinate system calibration are performed on multi-source data of land reserve areas to obtain a standardized dataset; Land reserve tiers are divided based on standardized datasets to obtain primary and secondary reserve data. Based on a unified spatiotemporal benchmark, data from primary and secondary reserve warehouses are dynamically correlated to obtain time-series correlation data for reserves. By integrating the ownership boundaries of the land reserve time-series data, a unified spatiotemporal database of land reserves can be obtained.

3. The method for multi-source integration and dynamic updating of land reserve data as described in claim 1, characterized in that, The process of extracting land ownership characteristics from a unified spatiotemporal database of land reserves to obtain land parcel ownership data, and then conducting a land compliance assessment on the land parcel ownership data to obtain land compliance assessment data, includes: Land ownership characteristics are extracted from the unified spatiotemporal database of land reserves to obtain land parcel ownership data; Land is classified based on land ownership data to obtain data on state-owned reserve land and collectively reserved land. Access the land compliance rules database, which includes land use control boundaries, constraints on historical buildings, ecological protection red line thresholds, and legal procedures for land expropriation; In accordance with land compliance rules, the data of state-owned reserve land parcels are assessed for compliance, thereby obtaining compliance assessment data of state-owned land parcels. In accordance with land compliance rules, the data of collectively reserved land parcels are used to conduct a compliance assessment of collective land parcels, thereby obtaining compliance assessment data of collective land parcels; The compliance assessment data of state-owned land parcels and collective land parcels are combined to resolve conflicts and obtain land compliance assessment data.

4. The method for multi-source integration and dynamic updating of land reserve data as described in claim 3, characterized in that, The process of conducting a compliance assessment of state-owned land reserve data based on land compliance rules to obtain state-owned land compliance assessment data includes: The data of state-owned reserve land parcels are divided into planning stages according to land compliance rules, thereby obtaining data on land parcels awaiting approval and land parcels that have been approved. Based on land compliance rules, the land use matching degree of approved land parcels is calculated to obtain compliant land parcel data and conflicting land parcel data; Compliance trajectory visualization is performed on compliant land parcel data and conflict land parcel data respectively to obtain compliance trajectory map sets and conflict trajectory map sets; Based on the compliance trajectory map set, legal matching statistics are performed to obtain compliance benchmark map data for state-owned land parcels; Based on the compliance benchmark map data of state-owned land parcels, source analysis of violations is performed on the conflict trajectory map set to obtain conflict land parcel assessment data; based on the compliance benchmark map data of state-owned land parcels, compliance index is calculated on the compliance trajectory map set to obtain compliant land parcel assessment data. The compliant land assessment data and the conflict land assessment data are weighted and merged to obtain the compliant assessment data of state-owned land.

5. The method for multi-source integration and dynamic updating of land reserve data as described in claim 4, characterized in that, The process of conducting a compliance assessment of collectively reserved land parcels based on land compliance rules to obtain collective land parcel compliance assessment data includes: Based on land compliance rules, the data of collective reserved land parcels are further subdivided by land nature to obtain data on collective commercial construction land, homestead land, and unused collective land. For data on collectively owned commercial construction land, the compliance of collective land transactions is verified, including checking the land transfer filing status, the approval rate of collective members' votes, and the publicity period of the income distribution plan, thereby obtaining compliance verification data for commercial land. For homestead data, the verification of homestead eligibility rights and use rights is carried out separately, including verifying the compliance of one homestead per household, records of penalties for exceeding the area occupied, and the status of historical ownership disputes, so as to obtain homestead ownership compliance data; For data on unused collective land, a dual verification of ecological protection and arable land protection is conducted, including comparing data on permanent basic farmland protection zones, historical industrial and mining land registration data, and soil pollution risk control lists, in order to obtain environmental compliance data for unused land. Based on compliance verification data for commercial land, compliance data for homestead ownership, and compliance data for the environment of unused land, a compliance weight matrix for collective land is constructed. The weight for commercial land focuses on procedural legality, the weight for homestead emphasizes the protection of rights and interests, and the weight for unused land strengthens ecological constraints. The three types of data are weighted and fused using a weight matrix, and data from the collective member objection feedback channel is injected for dynamic correction, thereby obtaining compliance assessment data for collective land parcels.

6. The method for multi-source integration and dynamic updating of land reserve data as described in claim 1, characterized in that, The process involves acquiring policy change data for land reserve areas and dynamically analyzing planning constraints based on this data to obtain policy constraint data. Regional land idling risk assessment is conducted based on a unified spatiotemporal database of land reserves and policy constraint data to obtain regional land idling risk data, including: Obtain data on policy changes in land reserve areas; Policy keywords and expiration dates are extracted from data on changes in land reserve policies, thereby obtaining structured policy data. The planning adjustment range is calculated from the structured policy data to obtain the planning adjustment data; Constraint type clustering is performed on the planning adjustment data to obtain rigid constraint data and flexible constraint data; Structured policy data is dynamically analyzed based on rigid constraint data and flexible constraint data to obtain policy constraint data; Based on the unified spatiotemporal database of land reserves and policy constraint data, combined with land transaction cycle data, idle land risk modeling is performed to obtain regional land idle risk data.

7. The method for multi-source integration and dynamic updating of land reserve data as described in claim 6, characterized in that, The method involves modeling the land idling risk based on a unified spatiotemporal database of land reserves and policy constraint data, combined with land transaction cycle data, to obtain regional land idling risk data, including: Constructing a land transaction map based on a unified spatiotemporal database of land reserves; The constraint intensity of policy constraint data is quantified to obtain a constraint intensity index; By aligning the land transaction map with the constraint intensity index in time and space, we can obtain land-policy correlation data. A land vacancy risk prediction model is constructed based on land parcel-policy correlation data and historical land vacancy records. By using an idle risk prediction model to calculate the risk probability of land parcel-policy correlation data, regional land idle risk data can be obtained.

8. The method for multi-source integration and dynamic updating of land reserve data as described in claim 1, characterized in that, The process involves predicting land value fluctuations from land compliance assessment data to obtain predicted land value data; then, based on regional land idling risk data, a weighted correction for the impact of idling is applied to the predicted land value data to obtain corrected land value data, which is then uploaded to the land reserve management platform to perform the task of marking abnormal value blocks. This includes: Land value fluctuations are predicted by analyzing land compliance assessment data, thereby obtaining land value prediction data. Based on land compliance assessment data and land compliance rules, policy change scenario simulations are conducted to obtain policy scenario simulation data. Marginal impact factors are then extracted from the policy scenario simulation data to obtain idleness impact factors. Risk-value coupling analysis is performed on regional land idling risk data and land value prediction data to obtain coupled anomaly data; A value correction model is constructed based on idle influencing factors and coupled abnormal data; The coupled abnormal data are dynamically weighted and corrected according to the value correction model to obtain corrected land value data, which is then uploaded to the land reserve management platform to perform the task of marking abnormal value blocks.

9. The method for multi-source integration and dynamic updating of land reserve data as described in claim 1, characterized in that, The process involves generating inefficient land revitalization strategies based on regional land idling risk data and policy constraint data, thereby obtaining inefficient land revitalization strategies, and uploading them to the land reserve management platform to execute dynamic land reserve optimization tasks, including: Based on regional land idling risk data, policy constraint data is clustered to identify the causes of idling, thereby obtaining causal classification data. Development potential features are extracted from the unified spatiotemporal database of land reserves to obtain land parcel development potential data; Based on the causal classification data and the land development potential data, revitalization strategies are matched to obtain a preliminary set of revitalization strategies; The initial set of revitalization strategies is verified for economic feasibility and policy compliance to obtain strategies for revitalizing inefficient land use. These strategies are then uploaded to the land reserve management platform to execute the dynamic optimization of land reserves.

10. A multi-source integration and dynamic updating system for land reserve data, characterized in that, The system includes: a memory, a processor, and a land reserve data multi-source integration and dynamic update program stored in the memory and executable on the processor, wherein the land reserve data multi-source integration and dynamic update program is configured to implement the steps of the land reserve data multi-source integration and dynamic update method as described in any one of claims 1 to 9.

Images