Data mining evaluation system for urban and rural planning and construction land

Abstract

The present invention discloses a data mining and evaluation system for urban and rural planning and construction land, and relates to the technical field of wind power data processing. The system comprises: a multidimensional data acquisition module for acquiring multidimensional data of target planning land. An evaluation relationship establishment module for respectively establishing evaluation relationships between multidimensional data and the rationality of planned land. A feature node mining module for taking each dimensional data as a feature node, mining its evaluation correlation, and establishing node connection edges. An evaluation interaction network construction module for fitting and connecting feature nodes and connection edges using the evaluation relationship to construct an evaluation interaction network. Feature segmentation is performed according to the evaluation interaction network to obtain a multi-layer structure of evaluation features, and hierarchical feature aggregation is performed based on this to generate an aggregated evaluation value, which is used for an aggregated evaluation module for evaluating planning rationality. Thereby achieving the technical effect of improving the depth and systematicity of data mining and improving the accuracy of evaluation.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a data mining and evaluation system for urban and rural planning and construction land. Background Art

[0002] In urban and rural planning, reasonable construction land assessment is an important link to ensure the efficient use of land resources. With the acceleration of urbanization, the complexity and diversity of planned land make it difficult for traditional assessment methods to meet actual needs. Therefore, with the help of data mining technology, multi-dimensional information is comprehensively analyzed. The current urban and rural planning assessment mainly relies on empirical judgment and simple data analysis, lacking in-depth mining and correlation analysis of multi-dimensional data. Although there are some GIS (geographic information system)-based tools for spatial data analysis, they are still insufficient in multi-dimensional data fusion and feature mining, and there are technical problems such as data islands, insufficient analysis and mining depth, and affecting the accuracy of assessment. Summary of the invention

[0003] The present invention provides a data mining and evaluation system for urban and rural planning and construction land to solve the technical problems of data islands, insufficient analysis and mining depth, and impact on evaluation accuracy in the prior art, and achieve the technical effect of improving data mining depth and systematicity and improving evaluation accuracy.

[0004] The data mining and evaluation system for urban and rural planning and construction land provided by the present invention comprises:

[0005] A multidimensional data acquisition module is used to acquire multidimensional data of the target planned land, wherein the multidimensional data includes basic information of the planned land, basic information of the surrounding land, building attributes and terrain information within a preset range, and regional environmental information.

[0006] An evaluation relationship establishment module is used to respectively establish the evaluation relationship between the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within a preset range, the regional environmental information and the rationality of the planned land.

[0007] The feature node mining module is used to use each dimension data as a feature node, mine the evaluation correlation between each feature node, and establish node connection edges.

[0008] An evaluation interaction network construction module is used to fit and connect the feature nodes and node connection edges using the evaluation relationship to construct an evaluation interaction network. An aggregation evaluation module is used to perform multi-level feature segmentation according to the evaluation interaction network to obtain a multi-level structure of evaluation features, perform hierarchical feature aggregation based on the multi-level structure of evaluation features, and obtain an aggregated evaluation value, which is used to evaluate the planning rationality of the target planning land.

[0009] In a feasible implementation, the evaluation relationship between the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within the preset range, the regional environmental information and the rationality of the planned land is established respectively, and the execution steps include:

[0010] Analyze the influencing factors of the rationality of the target planned land use, including the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within the preset range, and the regional environmental information.

[0011] Taking the influencing factors as independent variables and the impact on the target planned land as the dependent variable, historical monitoring data are used to perform relationship fitting to determine the influence relationship of each influencing factor.

[0012] The fitting coefficient is configured according to the proportion of the characteristic content of the influencing factors in the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within a preset range, and the regional environment information.

[0013] The fitting coefficient is used to integrate the influence relationship of the various influencing factors according to the mapping relationship, and an evaluation relationship between the basic information of planned land, basic information of surrounding land, building attributes and terrain information within a preset range, regional environmental information and the rationality of planned land is established.

[0014] In a feasible implementation, the factors affecting the rationality of the target planned land use are analyzed by analyzing the basic information of the planned land use, the basic information of the surrounding land use, the building attributes and terrain information within a preset range, and the regional environmental information. The execution steps include:

[0015] Through binary clustering of the assessment cases of planned construction land, successful cases and failed cases were obtained.

[0016] Evaluation factors are extracted based on the successful cases to obtain positive evaluation factors.

[0017] Abnormal events are extracted based on the failure cases to obtain reverse evaluation factors and abnormal event factors.

[0018] The positive evaluation factors, negative evaluation factors and abnormal event factors are used to obtain factor attribute characteristics.

[0019] The factor attribute characteristics are used to match and analyze the planned land basic information, surrounding land basic information, building attributes and terrain information within a preset range, and regional environmental information to obtain the influencing factors.

[0020] In a feasible implementation, the positive evaluation factor, the negative evaluation factor and the abnormal event factor are used to obtain the factor attribute characteristics, and the execution steps include:

[0021] The reverse evaluation factor and the abnormal event factor are cross-validated using the positive evaluation factor to obtain the repetition factor and the difference factor between the reverse evaluation factor and the abnormal event factor and the positive evaluation factor.

[0022] The difference factors in the positive evaluation factors are extracted as pure positive factors. The overlapping factors in the positive evaluation factors are obtained as pseudo positive factors.

[0023] The pseudo-positive factor is subjected to attribute analysis to obtain positive attributes and reverse attributes.

[0024] The reverse factor attribute is analyzed for the difference factor between the reverse evaluation factor and the abnormal event factor to obtain the reverse factor attribute. The reverse factor attribute feature is obtained according to the reverse attribute of the reverse factor attribute and the reverse attribute of the pseudo positive factor.

[0025] The pure positive factor is subjected to attribute analysis to obtain the positive factor attribute. The positive factor attribute characteristics are obtained according to the positive factor attribute and the positive attribute of the pseudo positive factor.

[0026] The factor attribute characteristics are obtained according to the positive factor attribute characteristics and the reverse factor attribute characteristics.

[0027] In a feasible implementation, the influencing factors are used as independent variables, the influence on the target planned land is used as the dependent variable, and the historical monitoring data is used to perform relationship fitting to determine the influence relationship of each influencing factor. The execution steps include:

[0028] The attribute characteristics of the influencing factors are identified based on the factor attribute characteristics.

[0029] Attributes are labeled by influencing factors according to attribute characteristics, and attribute labeling includes forward labeling and reverse labeling.

[0030] The fitting relationship attribute is determined by utilizing the attribute annotation, wherein the fitting relationship attribute annotated in the positive direction is a positive correlation, and the fitting relationship attribute annotated in the reverse direction is a negative correlation.

[0031] The relationship fitting is performed through the historical monitoring data, and the influence relationship of each influencing factor is obtained by combining the fitting relationship attributes.

[0032] In a feasible implementation, each dimension data is used as a feature node, the evaluation correlation between the feature nodes is mined, and the node connection edge is established. The execution steps include:

[0033] The dimensional data are taken as feature nodes, the influencing factors are taken as sub-nodes, node connection edges are established between the feature nodes and the sub-nodes, and the influence relationships of the influencing factors are fitted into the node connection edges.

[0034] The direct correlation between sub-nodes is identified through the correlation analysis operator.

[0035] Based on the influencing factors, time series data collection and prediction are performed, and time series correlation mining is performed through a deep mining model according to the time series prediction data to obtain indirect correlation.

[0036] According to the direct correlation and the indirect correlation, the node connection edge is configured, wherein the weight of the connection edge represents the strength of the correlation.

[0037] In a feasible implementation, multi-level feature segmentation is performed according to the evaluation interaction network to obtain a multi-level structure of evaluation features, and the execution steps include:

[0038] Based on the evaluation interaction network, the temporal influence and the strength of association of the association features are identified.

[0039] The temporal segmentation is performed according to the temporal influence of the correlation features to obtain a temporal segmentation hierarchical relationship.

[0040] In the time series segmentation level relationship, clustering is performed according to the correlation strength to obtain a correlation segmentation level.

[0041] In a feasible implementation, hierarchical feature aggregation is performed based on the multi-layer structure of the evaluation feature to obtain an aggregated evaluation value, and the execution steps include:

[0042] According to the relevance strength of each level in the relevance segmentation level, weighted aggregation is performed from the minimum clustering strength layer to the maximum clustering strength layer to obtain a strength aggregation result.

[0043] According to the intensity aggregation result, the evaluation value of the time series segmentation node is calculated.

[0044] Complete the evaluation values ​​of all time series segmentation nodes in sequence.

[0045] The evaluation weight of the time series segmentation node is configured, and the evaluation value of the time series segmentation hierarchical relationship is aggregated according to the time series relationship based on the evaluation weight to obtain the aggregated evaluation value.

[0046] The present invention discloses a data mining and evaluation system for urban and rural planning and construction land, including: a multidimensional data acquisition module, which is responsible for acquiring multidimensional data of the target planning land, including basic information of the planning land, basic information of the surrounding land, building attributes and terrain information within a preset range, and regional environmental information. An evaluation relationship establishment module is used to respectively construct the evaluation relationship between the basic information of the planning land, the basic information of the surrounding land, the building attributes and terrain information, the regional environmental information and the rationality of the planning land. A feature node mining module is used to regard each dimensional data as a feature node, mine the evaluation correlation between these feature nodes, and establish node connection edges. An evaluation interaction network construction module is used to fit and connect feature nodes and node connection edges using evaluation relationships, thereby constructing an evaluation interaction network. An aggregation evaluation module is used to perform multi-level feature segmentation according to the evaluation interaction network, obtain a multi-layer structure of evaluation features, and perform hierarchical feature aggregation based on this structure, and finally obtain an aggregate evaluation value, which is used to evaluate the rationality of the target planning land. The data mining and evaluation system for urban and rural planning and construction land disclosed in the present invention solves the technical problems of data islands, insufficient analysis and mining depth, and impact on evaluation accuracy, and achieves the technical effect of improving the depth and systematicness of data mining and improving evaluation accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a structural schematic diagram of the data mining and evaluation system for urban and rural planning and construction land of the present invention;

[0048] Figure 2 It is a schematic diagram of the process of establishing an evaluation relationship in the data mining evaluation system for urban and rural planning and construction land of the present invention.

[0049] Explanation of the reference numerals: multi-dimensional data acquisition module 11, evaluation relationship establishment module 12, characteristic node mining module 13, evaluation interaction network construction module 14, aggregation evaluation module 15. DETAILED DESCRIPTION

[0050] The technical solution provided in the embodiments of the present invention is to solve the technical problems of data islands, insufficient analysis and mining depth, and impact on evaluation accuracy in the prior art. The overall idea adopted is as follows:

[0051] First, the multidimensional data of the target planning land is obtained through the multidimensional data acquisition module, including the basic information of the planning land, the basic information of the surrounding land, the building attributes and terrain information within the preset range, and the regional environmental information. Then, the evaluation relationship establishment module constructs the evaluation relationship between the basic information of the planning land, the basic information of the surrounding land, the building attributes and terrain information, the regional environmental information and the rationality of the planning land. Next, the feature node mining module regards each dimensional data as a feature node, mines the evaluation correlation between these feature nodes, and establishes node connection edges. Then, through the evaluation interaction network construction module, the evaluation relationship is used to fit and connect the feature nodes and node connection edges, thereby constructing an evaluation interaction network. Finally, the aggregation evaluation module performs multi-level feature segmentation according to the evaluation interaction network to obtain the multi-level structure of the evaluation feature, and performs hierarchical feature aggregation based on this structure to obtain the aggregated evaluation value, which is used to evaluate the rationality of the target planning land.

[0052] The above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods of the specification to better understand the above technical solution. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments used only to explain the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, it should be noted that, for the convenience of description, only the parts related to the present invention are shown in the drawings, rather than all of them. Example

[0053] Figure 1 The flowchart of the data mining and evaluation system for urban and rural planning and construction land of the present invention is shown in FIG. 1 , wherein the system includes:

[0054] The multi-dimensional data acquisition module 11 is used to acquire multi-dimensional data of the target planned land, wherein the multi-dimensional data includes basic information of the planned land, basic information of surrounding land, building attributes and terrain information within a preset range, and regional environment information.

[0055] Specifically, the multidimensional data of the target planned land includes basic information of the planned land, basic information of the surrounding land, building attributes and terrain information within a preset range, and regional environmental information. The multidimensional data can comprehensively reflect the basic information of the target planned land, help analyze the relationship between the construction land and the surrounding environment, and thus provide a scientific basis for urban planning and resource allocation. Exemplarily, the above multidimensional data is obtained through remote sensing technology, geographic information system, field survey or access to data archives.

[0056] Specifically, the basic information of planned land use includes: the nature of land use (including residential, commercial, industrial and other categories); land area, that is, the specific area of ​​the land; the nature of land use rights, such as transfer, allocation and other types; planning indicators, such as volume ratio, building density, green space ratio, etc.;

[0057] Specifically, the basic information of surrounding land use includes: surrounding land use, that is, the use and attributes of adjacent plots, such as whether they are commercial or residential areas. The current land use status refers to the actual use status of the current plot and its changing trend.

[0058] Specifically, the building attributes within the preset range include: building type, building height and scale, building usage (such as utilization rate and activity). The terrain information within the preset range includes: topography, water system distribution, etc. Among them, the preset range is determined based on the land use nature of the target planned land, and the protection or control range of planned land with different land use natures is not the same.

[0059] Specifically, regional environmental information includes: ecological environment (distribution of surrounding nature reserves, green spaces, wetlands, etc.); distribution of environmentally sensitive areas, used for impact assessment of specific areas such as schools and hospitals to quantify the potential impact on planned land; transportation network information, etc.

[0060] The evaluation relationship establishment module 12 is used to respectively establish the evaluation relationship between the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within a preset range, the regional environmental information and the rationality of the planned land.

[0061] Specifically, through the evaluation relationship establishment module, the direct evaluation relationship between the basic information of planned land, the basic information of surrounding land, the building attributes and terrain information within the preset range, the regional environmental information and the rationality of planned land is established respectively. Among them, the evaluation relationship can be considered as an isolated mapping between the basic information of planned land, the basic information of surrounding land, the building attributes and terrain information within the preset range, the regional environmental information and the rationality of planned land.

[0062] In some embodiments, Figure 2 As shown, the evaluation relationship between the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within the preset range, the regional environmental information and the rationality of the planned land is established respectively, and the execution steps include:

[0063] Analyze the influencing factors of the basic information of planned land, the basic information of surrounding land, the building attributes and terrain information within a preset range, and the regional environmental information on the rationality of the target planned land; take the influencing factors as independent variables and the influence on the target planned land as dependent variables, use historical monitoring data to perform relationship fitting, and determine the influence relationship of each influencing factor; configure fitting coefficients according to the proportion of the characteristic content of the influencing factors in the basic information of planned land, the basic information of surrounding land, the building attributes and terrain information within a preset range, and the regional environmental information; use the fitting coefficients to integrate the influence relationships of the influencing factors according to the mapping relationship, and establish an evaluation relationship between the basic information of planned land, the basic information of surrounding land, the building attributes and terrain information within a preset range, the regional environmental information and the rationality of planned land.

[0064] Specifically, firstly, based on the historical planning land assessment cases, the specific influencing factors that affect the rationality of the target planning land in the basic information of the planning land, the basic information of the surrounding land, the building attributes and terrain information within the preset range, and the regional environmental information are determined. Exemplarily, the above-mentioned influencing factors include multiple specific information indicators in the multi-dimensional information, and each influencing factor is marked with a corresponding influencing direction (i.e., positive or negative).

[0065] Specifically, through appropriate statistical or machine learning methods, such as linear regression, logistic regression, random forest, etc., the influencing factors are set as independent variables, and the rationality indicators of the target planned land (such as evaluation scores, proportion of compliance with standards, etc.) are set as dependent variables. Taking historical monitoring data as sample data, the regression coefficients or weights between the independent variables and the dependent variables are fitted as the influence relationship of each influencing factor.

[0066] Specifically, the characteristic content ratio of each influencing factor in the data set of relevant planning land basic information, surrounding land basic information, building attributes and terrain information within a preset range, and regional environmental information is calculated. The characteristic content ratio is used to reflect the importance of the influencing factor. In other words, the higher the characteristic content ratio of the influencing factor, the higher the expected impact effect. Then, the obtained multiple characteristic content ratios are normalized respectively, and the normalized results are configured as fitting coefficients. The fitting coefficients configured by the above process can better reflect the situation of the influencing factor in actual application, thereby improving the accuracy of the evaluation.

[0067] Specifically, the fitting relationship of each influencing factor and the configured fitting coefficient are integrated, and a corresponding comprehensive evaluation model is formed through the weighted fitting method, which serves as the evaluation relationship between the basic information of planned land, the basic information of surrounding land, the building attributes and terrain information within a preset range, the regional environmental information and the rationality of planned land.

[0068] In some implementations, the steps of analyzing the factors affecting the rationality of the target planned land use by the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within a preset range, and the regional environmental information include:

[0069] Through binary clustering of the assessment cases of planned construction land, successful cases and failed cases are obtained; based on the successful cases, assessment factors are extracted to obtain positive assessment factors; based on the failed cases, abnormal events are extracted to obtain reverse assessment factors and abnormal event factors; using the positive assessment factors, reverse assessment factors and abnormal event factors, factor attribute characteristics are obtained; using the factor attribute characteristics, the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within a preset range, and the regional environment information are matched and analyzed to obtain the influencing factors.

[0070] Specifically, firstly, multiple assessment cases of planned construction land are obtained. Preferably, the assessment cases have the same or similar attribute characteristics, such as industrial land, similar land quality or basic characteristics, similar climate or surrounding environment, etc. Then, according to the assessment effectiveness or accuracy of the assessment cases, the multiple assessment cases are divided into successful cases and failed cases.

[0071] Specifically, through statistical methods, we analyze and extract common characteristics from successful cases as positive evaluation factors that help improve the rationality of planned land use, such as good transportation connections, sufficient public service facilities, etc.

[0072] Specifically, based on the same principle of extracting positive evaluation factors, the characteristics of failure cases are analyzed, the common factors leading to failure are found, and the factors that have a negative impact on the rationality of planned land use are extracted to form reverse evaluation factors; at the same time, abnormal events that occur in failure cases (such as environmental pollution, etc.) are recorded and classified as abnormal event factors.

[0073] Specifically, all extracted positive evaluation factors, negative evaluation factors and abnormal event factor characteristics are analyzed, the direction of action of each evaluation factor is verified and determined, and the output is the factor attribute characteristics. In other words, the evaluation factors and category labels of the evaluation factors obtained based on the above statistical analysis method are preliminary classification results. The specific effect characteristics of different evaluation factors on the rationality of the target planning land may be different from the above-determined category labels (positive, negative, abnormal), and need to be redefined through factor attribute characteristics to ensure the efficiency and accuracy of the subsequent generation of influencing factors.

[0074] Specifically, the determined factor attribute characteristics are used as index targets, and the basic information of planned land, the basic information of surrounding land, the building attributes and terrain information within the preset range, and the regional environmental information are traversed for matching and analysis, so as to obtain multiple influencing factors related to the rationality of the target planned land in the multi-dimensional information of the target scene. Through the above matching and analysis process, it helps to ensure the complete acquisition of the elements related to the rationality of the target planned land, while avoiding the interference of invalid indicator data and reducing the cost of subsequent data processing and evaluation processing.

[0075] Furthermore, the positive evaluation factor, the negative evaluation factor and the abnormal event factor are used to obtain the factor attribute characteristics, and the execution steps include:

[0076] The positive evaluation factor is used to cross-validate the reverse evaluation factor and the abnormal event factor respectively, and the repeated factors and difference factors of the reverse evaluation factor, the abnormal event factor and the positive evaluation factor are obtained; the difference factor in the positive evaluation factor is extracted as a pure positive factor; the overlapping factor in the positive evaluation factor is obtained as a pseudo positive factor; the pseudo positive factor is attributed and analyzed to obtain positive attributes and reverse attributes; the difference factor of the reverse evaluation factor and the abnormal event factor is attributed and analyzed to obtain reverse factor attributes; the reverse factor attribute characteristics are obtained according to the reverse factor attributes and the reverse attributes of the pseudo positive factor; the pure positive factor is attributed and analyzed to obtain positive factor attributes; the positive factor attribute characteristics are obtained according to the positive factor attributes and the positive attributes of the pseudo positive factor; the factor attribute characteristics are obtained according to the positive factor attribute characteristics and the reverse factor attribute characteristics.

[0077] Specifically, first, the positive evaluation factor is cross-validated with the negative evaluation factor and the abnormal event factor respectively, so as to identify repeated factors (i.e., indicator dimensions that exist in both the positive and another factor) and difference factors (indicator dimensions that only exist in a single type of factor).

[0078] Furthermore, the above-mentioned difference factors are extracted from the positive evaluation factors and defined as pure positive factors, which have an absolute positive contribution to the rationality of the target planned land use; the above-mentioned overlapping factors are extracted from the positive evaluation factors and defined as pseudo-positive factors, which are factors in the positive factors that have an incomplete positive contribution.

[0079] Specifically, a detailed attribute analysis is performed on the pseudo-positive factors to obtain the positive attributes (the scope of action that has a positive impact on rationality) and the negative attributes (the scope of action that brings negative impact) of the pseudo-positive factors. For example, for the influencing factors related to the field of transportation infrastructure, the influencing attributes (negative attributes) of the construction land for recuperation purposes are different from the influencing attributes (positive attributes) of the construction land for industrial or logistics purposes.

[0080] Specifically, the attribute analysis of the difference factor between the reverse evaluation factor and the abnormal event factor is performed to determine the reverse factor attribute of the difference factor. Then, the reverse factor attribute is combined with the reverse attribute of the pseudo-positive factor to form the reverse factor attribute feature, providing a detailed description of the negative impact of the influencing factor.

[0081] Specifically, the attributes of pure positive factors are analyzed to obtain positive factor attributes, which include the degree of positive impact (such as mild, ordinary, strong, extremely strong, etc.). Then, combined with the positive attributes of pseudo-positive factors, the positive factor attribute characteristics are formed to clarify all the influencing factors that produce positive effects and the corresponding range conditions.

[0082] In some implementations, the influencing factors are used as independent variables, the influence on the target planned land is used as the dependent variable, and historical monitoring data are used to perform relationship fitting to determine the influence relationship of each influencing factor. The execution steps include:

[0083] Identify the attribute characteristics of the influencing factors based on the factor attribute characteristics; perform attribute annotation on the influencing factors according to the attribute characteristics, wherein the attribute annotation includes forward annotation and reverse annotation; determine the fitting relationship attributes using the attribute annotation, wherein the fitting relationship attributes of the forward annotation are positive correlation, and the fitting relationship attributes of the reverse annotation are negative correlation; perform relationship fitting through the historical monitoring data, and obtain the influence relationship of each influencing factor in combination with the fitting relationship attributes.

[0084] Specifically, first, based on the factor attribute characteristics obtained in the previous steps and according to the basic information of the target planning land, the specific attribute characteristics of each influencing factor are identified, that is, the direction of influence of the influencing factor on the rationality of the target planning land is determined, and positive and negative annotations are performed accordingly, and the influencing factors are divided into positive influencing factors and negative influencing factors.

[0085] Specifically, according to the attribute annotation of the influencing factor, the fitting relationship attribute is determined. If the factor is labeled as positive, it means that it is positively correlated with the impact of the target planning land (it has a positive impact on the target planning land). If the factor is labeled as negative, it means that it is negatively correlated with the impact of the target planning land (it has a negative impact on the target planning land).

[0086] Specifically, based on historical monitoring data, the fitting direction is initialized with the fitting relationship attributes, and a suitable statistical or machine learning model (such as linear regression, decision tree, etc.) is selected to fit the relationship between the influencing factors and the impact of the target planning land use, and the degree and direction of influence between each influencing factor and the impact of the target planning land use are obtained, and the output is the influence relationship of each influencing factor.

[0087] Through the above steps, a quantitative relationship between the influencing factors and the rationality of the target planning land use can be established, thereby providing data support and theoretical basis for subsequent analysis and ensuring that various influencing factors are effectively considered in the evaluation process.

[0088] The feature node mining module 13 is used to use each dimension data as a feature node, mine the evaluation correlation between each feature node, and establish node connection edges.

[0089] In some embodiments, each dimension data is used as a feature node, the evaluation correlation between each feature node is mined, and the node connection edge is established. The execution steps include:

[0090] The data of each dimension is taken as a feature node, and the influencing factor is taken as a sub-node, and a node connection edge between the feature node and the sub-node is established, and the influence relationship of each influencing factor is fitted into the node connection edge; the direct correlation between the sub-nodes is identified through a correlation analysis operator; time series data is collected and predicted based on the influencing factor, and time series correlation mining is performed through a deep mining model according to the time series prediction data to obtain indirect correlation; sub-node connection edges are configured according to the direct correlation and indirect correlation; wherein the weight of the connection edge represents the strength of the correlation.

[0091] Specifically, first, each dimension data is regarded as a feature node, and according to the relationship between the influencing factor and each dimension data, the influencing factor is used as a sub-node under the feature node. Then, based on the influence relationship of each influencing factor, a node connection edge is established between the feature node and multiple sub-nodes. The node connection edge is the relationship edge between the influencing factor and each dimension data, marked with the relationship strength information (i.e., the influence relationship).

[0092] Specifically, the Pearson correlation coefficient, the Spearman rank correlation coefficient or the Kendall rank correlation coefficient is used as a correlation analysis operator to calculate the direct correlation between the feature node and the sub-node.

[0093] Specifically, time series data is collected based on influencing factors, and a deep mining model is used to mine time series correlations to obtain indirect correlations. The deep mining model is built based on a convolutional neural network (CNN) or a temporal convolutional network (TCN) architecture to obtain long-term dependencies. The indirect correlations obtained through the deep mining model reflect the changing characteristics of the relationship between feature nodes and sub-nodes in the time dimension.

[0094] Specifically, according to the direct correlation and indirect correlation, the connection edges between the sub-nodes are configured, and the weight of the connection edge represents the strength of the correlation, which is used to reflect the closeness of the relationship between the data nodes. For example, if the indirect correlation is characterized as decreasing over time, the direct correlation needs to be appropriately adjusted to reduce the weight of the sub-node connection edge to meet the situation in the actual scenario; if the indirect correlation is characterized as a periodic change, the sub-node connection edge is defined based on the periodic change characteristics combined with the direct correlation (such as configuring a weight function or calculating an equivalent correlation).

[0095] The above-mentioned connection edge configuration strategy combining direct correlation and indirect correlation fully considers the changes or fluctuations of the influence relationship of influencing factors in the time dimension, which helps to improve the reliability and foresight of land use evaluation.

[0096] The evaluation interaction network construction module 14 is used to fit and connect the feature nodes and node connection edges using the evaluation relationship to construct an evaluation interaction network.

[0097] Specifically, the evaluation relationship can be considered as the upper-level weight of the evaluation interaction network, that is, the weight of the relationship edge between the feature node and the planning land rationality result. In other words, the connection edge between each feature node and its corresponding planning land rationality is assigned a corresponding weight based on the evaluation relationship.

[0098] Through the fitting connection of the characteristic nodes and node connection edges based on the evaluation relationship, an evaluation path from each sub-node from top to bottom to multiple characteristic nodes and finally pointing to the rationality of planned land use is formed as an evaluation interaction network. The relationship between characteristic nodes can be quantified and the impact of various factors on the rationality of planned land use can be intuitively reflected.

[0099] The aggregation evaluation module 15 is used to perform multi-level feature segmentation according to the evaluation interaction network to obtain a multi-level structure of evaluation features, perform hierarchical feature aggregation based on the multi-level structure of evaluation features to obtain an aggregation evaluation value, and the aggregation evaluation value is used to evaluate the planning rationality of the target planning land.

[0100] In some embodiments, multi-level feature segmentation is performed according to the evaluation interaction network to obtain a multi-level structure of evaluation features, and the execution steps include:

[0101] According to the evaluation interaction network, the temporal influence and the correlation strength of the correlation features are identified; temporal segmentation is performed according to the temporal influence of the correlation features to obtain a temporal segmentation hierarchical relationship; in the temporal segmentation hierarchical relationship, clustering is performed according to the correlation strength to obtain a correlation segmentation hierarchy.

[0102] Specifically, first, based on the evaluation interaction network, the correlation features between feature nodes are analyzed, and a list of identified correlation features is output, which contains the temporal impact and strength of each feature. Then, according to the temporal impact of the identified correlation features, the features are segmented to form a temporal segmentation hierarchy to ensure the effectiveness and consistency of the relevant features in the time dimension. Next, in the established temporal segmentation hierarchy, clustering algorithms (such as K-means or hierarchical clustering) are used to perform cluster analysis on the features based on the strength of the correlation, and features with high similarity are divided into the same group to obtain the correlation segmentation hierarchy.

[0103] Through the above steps, a multi-layer structure of evaluation features was successfully constructed, and effective segmentation of features was achieved, which is helpful for further analysis of how features at different levels affect the rationality of planned land use.

[0104] In some embodiments, hierarchical feature aggregation is performed based on the multi-layer structure of the evaluation features to obtain an aggregated evaluation value, and the execution steps include:

[0105] According to the correlation strength of each level in the correlation segmentation hierarchy, weighted aggregation is performed from the minimum clustering strength layer to the maximum clustering strength layer to obtain a strength aggregation result; based on the strength aggregation result, the evaluation value of the time series segmentation node is calculated; the evaluation values ​​of all time series segmentation nodes are completed in sequence; the evaluation weights of the time series segmentation nodes are configured, and based on the evaluation weights, the evaluation values ​​of the time series segmentation hierarchy relationships are aggregated according to the time series relationship to obtain the aggregated evaluation value.

[0106] Specifically, a corresponding weight is assigned to each level according to the strength of association, and weighted aggregation is performed from the minimum clustering strength layer to the maximum clustering strength layer to obtain the strength aggregation result. Among them, the small clustering strength layer can be considered as the lowest indicator dimension. Through this layer-by-layer weighted aggregation strategy, the information of multiple indicators can be effectively integrated, improving the rationality of complex comprehensive evaluation and the accuracy of evaluation results.

[0107] For example, for the road traffic level, the minimum clustering strength layer may include road mileage, road grade distribution, traffic volume, travel speed, etc., while the upper clustering strength layer of the minimum clustering strength layer may correspond to the congestion index, average commuting time, etc. obtained by weighted aggregation of indicators in the minimum clustering strength layer.

[0108] Specifically, based on the aforementioned evaluation interaction network, the aggregation results are mapped to the evaluation values ​​of the time-series segmentation nodes, and the evaluation values ​​of multiple time-series segmentation nodes are obtained; then, different evaluation weights are assigned to each time-series segmentation node according to the importance of the time period, and the weighted aggregation of the evaluation values ​​of multiple time-series segmentation hierarchical relationships is performed according to the evaluation weights to obtain the final aggregated evaluation value. The above-mentioned evaluation value aggregation process combining time-series segmentation and weighted aggregation helps to capture the changing characteristics of data over time, thereby better understanding the dynamic impact of planned land and its surrounding environment, and providing an aggregated evaluation value with higher accuracy and reference.

[0109] In summary, the data mining and evaluation system for urban and rural planning and construction land provided by the present invention has the following technical effects:

[0110] The multidimensional data acquisition module obtains the multidimensional data of the target planning land, including the basic information of the planning land, the basic information of the surrounding land, the building attributes and terrain information within the preset range, and the regional environmental information. The evaluation relationship establishment module constructs the evaluation relationship between the basic information of the planning land, the basic information of the surrounding land, the building attributes and terrain information, the regional environmental information and the rationality of the planning land. The feature node mining module regards each dimensional data as a feature node, mines the evaluation correlation between these feature nodes, and establishes node connection edges. The evaluation interaction network construction module uses the evaluation relationship to fit and connect the feature nodes and node connection edges, thereby constructing an evaluation interaction network. The aggregation evaluation module performs multi-level feature segmentation according to the evaluation interaction network, obtains the multi-level structure of the evaluation feature, and performs hierarchical feature aggregation based on this structure, and finally obtains the aggregated evaluation value, which is used to evaluate the rationality of the target planning land. Thus, the technical effect of improving the depth and systematicity of data mining and improving the accuracy of evaluation is achieved.

[0111] It should be understood that the embodiments disclosed in the present invention and the above description can enable those skilled in the art to use the present invention to implement the present invention. At the same time, the present invention is not limited to the above-mentioned embodiments. It should be understood that those skilled in the art can still modify the technical solutions recorded in the above-mentioned embodiments, or replace some of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. The data mining and evaluation system for urban and rural planning and construction land is characterized by: The system comprises: A multi-dimensional data acquisition module, which is used to acquire multi-dimensional data of the target planned land, wherein the multi-dimensional data includes basic information of the planned land, basic information of surrounding land, building attributes and terrain information within a preset range, and regional environment information; An evaluation relationship establishment module, the evaluation relationship establishment module is used to respectively establish an evaluation relationship between the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within a preset range, the regional environmental information and the rationality of the planned land; A feature node mining module, wherein the feature node mining module is used to use each dimension data as a feature node, mine the evaluation correlation between each feature node, and establish node connection edges; An evaluation interaction network construction module, wherein the evaluation interaction network construction module is used to fit and connect the feature nodes and node connection edges using the evaluation relationship to construct an evaluation interaction network; An aggregation evaluation module, which is used to perform multi-level feature segmentation according to the evaluation interaction network to obtain a multi-level structure of evaluation features, and to perform hierarchical feature aggregation based on the multi-level structure of evaluation features to obtain an aggregated evaluation value, wherein the aggregated evaluation value is used to evaluate the planning rationality of the target planned land; The evaluation relationship between the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within the preset range, the regional environmental information and the rationality of the planned land is established respectively, and the execution steps include: Analyze the factors affecting the rationality of the target planned land use by the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within the preset range, and the regional environmental information; Taking the influencing factors as independent variables and the impact on the target planned land as dependent variables, historical monitoring data are used to perform relationship fitting to determine the impact relationship of each influencing factor; According to the proportion of the influencing factors in the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within the preset range, and the regional environment information, the fitting coefficient is configured; The fitting coefficient is used to integrate the influence relationship of the various influencing factors according to the mapping relationship, and an evaluation relationship between the basic information of planned land, basic information of surrounding land, building attributes and terrain information within a preset range, regional environmental information and the rationality of planned land is established.

2. The data mining and evaluation system for urban and rural planning and construction land according to claim 1, characterized in that: Analyze the factors affecting the rationality of the target planned land use by the basic information of the planned land, the basic information of the surrounding land, the building attributes and terrain information within a preset range, and the regional environmental information, and the execution steps include: Through binary clustering of the evaluation cases of planned construction land, successful cases and failed cases were obtained; Extract evaluation factors based on the successful cases to obtain positive evaluation factors; Extract abnormal events based on the failure cases to obtain reverse evaluation factors and abnormal event factors; Using the positive evaluation factors, negative evaluation factors and abnormal event factors, factor attribute characteristics are obtained; The factor attribute characteristics are used to match and analyze the planned land basic information, surrounding land basic information, building attributes and terrain information within a preset range, and regional environmental information to obtain the influencing factors.

3. The data mining and evaluation system for urban and rural planning and construction land according to claim 2, characterized in that: Using the positive evaluation factor, the negative evaluation factor and the abnormal event factor, the factor attribute characteristics are obtained, and the execution steps include: Using the positive evaluation factor to cross-validate the negative evaluation factor and the abnormal event factor, respectively, to obtain the repetition factor and the difference factor between the negative evaluation factor and the abnormal event factor and the positive evaluation factor; Extracting the difference factors in the positive evaluation factors as pure positive factors; obtaining the overlapping factors in the positive evaluation factors as pseudo positive factors; Performing attribute analysis on the pseudo positive factor to obtain positive attributes and reverse attributes; Performing attribute analysis on the difference factors of the reverse evaluation factor and the abnormal event factor to obtain reverse factor attributes; obtaining reverse factor attribute characteristics according to the reverse factor attributes and the reverse attributes of the pseudo positive factor; Performing attribute analysis on the pure positive factor to obtain positive factor attributes; obtaining positive factor attribute characteristics according to the positive factor attributes and the positive attributes of the pseudo positive factor; The factor attribute characteristics are obtained according to the positive factor attribute characteristics and the reverse factor attribute characteristics.

4. The data mining and evaluation system for urban and rural planning and construction land according to claim 3, characterized in that: Taking the influencing factors as independent variables and the influence on the target planned land as dependent variables, historical monitoring data are used to perform relationship fitting to determine the influence relationship of each influencing factor. The execution steps include: Identifying the attribute characteristics of the influencing factor based on the factor attribute characteristics; Attribute annotation is performed based on the influencing factors according to the attribute characteristics, wherein the attribute annotation includes forward annotation and reverse annotation; Determine the fitting relationship attribute by using the attribute annotation, wherein the fitting relationship attribute annotated in the positive direction is a positive correlation, and the fitting relationship attribute annotated in the reverse direction is a negative correlation; The relationship fitting is performed through the historical monitoring data, and the influence relationship of each influencing factor is obtained by combining the fitting relationship attributes.

5. The data mining and evaluation system for urban and rural planning and construction land according to claim 1, characterized in that: Take each dimension data as a feature node, mine the evaluation correlation between each feature node, and establish node connection edges. The execution steps include: Taking the dimensional data as feature nodes and the influencing factors as sub-nodes, establishing node connection edges between the feature nodes and the sub-nodes, and fitting the influence relationships of the influencing factors into the node connection edges; Identify the direct correlation between sub-nodes through correlation analysis operators; Based on the influencing factors, time series data is collected and predicted, and time series correlation mining is performed through a deep mining model according to the time series prediction data to obtain indirect correlation; According to the direct association and the indirect association, the node connection edges are configured; wherein the weight of the connection edge represents the strength of the association.

6. The data mining and evaluation system for urban and rural planning and construction land according to claim 5, characterized in that: Perform multi-level feature segmentation according to the evaluation interaction network to obtain a multi-level structure of evaluation features, and the execution steps include: identifying the temporal influence and the strength of association of the association features according to the evaluation interaction network; Performing time series segmentation according to the time series influence of the correlation features to obtain a time series segmentation hierarchical relationship; In the time series segmentation level relationship, clustering is performed according to the correlation strength to obtain a correlation segmentation level.

7. The data mining and evaluation system for urban and rural planning and construction land according to claim 6, characterized in that: Based on the multi-layer structure of the evaluation features, hierarchical feature aggregation is performed to obtain an aggregated evaluation value, and the execution steps include: According to the correlation strength of each level in the correlation segmentation level, weighted aggregation is performed from the minimum clustering strength layer to the maximum clustering strength layer to obtain a strength aggregation result; Calculating the evaluation value of the time series segmentation node according to the strength aggregation result; Complete the evaluation values ​​of all time series segmentation nodes in sequence; The evaluation weight of the time series segmentation node is configured, and the evaluation value of the time series segmentation hierarchical relationship is aggregated according to the time series relationship based on the evaluation weight to obtain the aggregated evaluation value.

Images