CN-122022882-A - Method for analyzing, studying and judging value of removed land block based on GIS and one-user one-file data

CN122022882ACN 122022882 ACN122022882 ACN 122022882ACN-122022882-A

Abstract

The invention relates to a method for analyzing and judging a migration land parcel value based on GIS and one-user one-file data, which is characterized by realizing the integrity and space consistency of a structured data set by collecting and standardizing one-user one-file attribute data and GIS space data, establishing accurate mapping of attributes and space entities by utilizing a space analysis technology, generating a composite analysis unit and checking the space attribution relation of the composite analysis unit, constructing a multi-dimensional land parcel evaluation model, calculating development value and implementation cost comprehensive indexes by adopting a feature normalization, dimension reduction and machine learning method, forming a semantic rule knowledge base by combining historical text mining, generating an optimal development suggestion based on semantic reasoning and similarity matching, outputting strategy combination and strategy adaptation scheme by comprehensively integrating position, weight and matching factors of a decision tree, and establishing a manual feedback closed loop to iteratively optimize decision accuracy.

Inventors

LIN TAO
CHEN HANG
WU XIAOJIAN
HU HESONG
SHEN HAO
LI XIN
LIU YADONG
Quan Ruiwen

Assignees

广州市建筑科学研究院集团有限公司
广州市市政集团有限公司房地产置业分公司
广州建设工程质量安全检测中心有限公司

Dates

Publication Date: 20260512
Application Date: 20260108

Claims (10)

1. The method for analyzing and judging the value of the removed land block based on GIS and one-user one-file data is characterized by comprising the following steps: S1, collecting first-household first-grade attribute data and GIS space data in an urban updating project, carrying out standardized cleaning on the first-household first-grade attribute data, and carrying out coordinate system first processing on the GIS space data; S2, establishing a mapping relation between one-user one-file data and GIS space data elements through a space connection tool; s3, constructing a multidimensional migration plot evaluation model, dividing a development value index calculation module and an implementation cost calculation module, and generating a plot comprehensive evaluation index by adopting a weighted summation method; s4, performing text mining on the historical city update project files, extracting decision rules of compensation standards, volume rate adjustment and placement schemes, and constructing a semantic rule knowledge base containing triples of land block features, business actions and implementation effects; S5, constructing a semantic reasoning engine comprising a feature mapping module, a rule matching module and a logic reasoning module, inputting the plot comprehensive evaluation index into the semantic reasoning engine, calculating a similarity retrieval optimal decision path between the current plot feature and the historical plot feature in the semantic rule knowledge base, and reasoning based on the mapping relation between the business action and the implementation effect in the triplet to generate a candidate suggestion set comprising a volume rate improvement application, a compensation standard adjustment and a placement scheme optimization; s6, based on the block space location, the weight complexity and the peripheral matched maturity characteristics, multidimensional decision tree analysis is performed on the candidate suggestion set, and development strategy combination with implementation priority ranking and a corresponding policy adaptation scheme are output.
2. The method for analyzing and judging the value of a removed land based on GIS and one-user one-file data according to claim 1, wherein the following step S6 further comprises: s7, generating a visual decision map, displaying a causal association network of an evaluation index, a plot characteristic and a recommendation strategy, marking supporting evidence sources of the recommendation strategies, and automatically generating an evaluation report abstract containing a value basis and an implementation path; And S8, establishing a manual feedback closed-loop mechanism, recording deviation reasons and updating decision weight parameters in a semantic rule knowledge base when the actual decision deviates from a system recommendation scheme, and optimizing the case matching accuracy of the semantic reasoning engine through online learning.
3. The method for analyzing and studying and judging the value of a removed land based on GIS and one-household one-file data according to claim 1, wherein the one-household one-file attribute data comprises property types, population structures and history compensation records.
4. The method for analyzing and judging the value of a removed land mass based on GIS and one-user one-file data according to claim 1, wherein the GIS space data comprises land mass boundary, volume rate and peripheral matching information.
5. The method for analyzing and judging the value of a removed land parcel based on GIS and one-user one-file data according to claim 4, wherein the space data acquisition interface module is adopted to realize the data capture of the land parcel boundary, volume rate and peripheral matched information vector layers to obtain the land parcel boundary vector layer, the volume rate vector layer and the peripheral matched information vector layer.
6. The method for analyzing and studying and judging the value of a removed land parcel based on GIS and one-user first-class data according to claim 5, wherein the geometric data and attribute table of the land parcel boundary vector layer, the volume rate vector layer and the surrounding matched information vector layer are structurally analyzed, an original space data set containing a coordinate point column and an attribute field is obtained, and the vector layer in the original space data set is uniformly converted into a target coordinate system based on EPSG standard coordinate system, so that the space data set under a uniform coordinate system is obtained.
7. The method for analyzing and studying and judging the value of the removed land based on GIS and one-user one-file data according to claim 2, wherein the manual feedback closed-loop mechanism performs category attribution on the condition that the actual decision deviates from the recommendation of the system, and the deviation type covers policy limitation, rights and interests complexity and social public opinion.
8. The method for analyzing and judging the value of a removed land based on GIS and one-user one-file data according to claim 1, wherein the step S4 specifically comprises: preprocessing unstructured text data in a history city update project file to obtain a structured text feature vector; performing semantic embedded coding on the structured text feature vector based on the BERT pre-training language model to generate a text semantic representation matrix; Identifying and extracting triple information fragments related to land feature, business action and implementation effect from a text semantic representation matrix by adopting a method based on combination of rule matching and dependency syntactic analysis so as to form a preliminary decision rule candidate set; Performing relationship classification and entity alignment processing on the preliminary decision rule candidate set, and performing consistency verification on the relationship in the triples based on the knowledge graph embedding model TransE; and storing the consistency checked triplet information into a graph database according to a unified semantic expression format, and constructing a semantic rule knowledge base containing land parcel features, business actions and implementation effects.
9. The method for analyzing and studying and judging the value of a removed land based on GIS and one-user one-file data according to claim 8, wherein the preprocessing of unstructured text data in the history city update project file comprises word segmentation, stop word removal, word stem extraction and named entity recognition.
10. The method for analyzing and judging the value of a removed land based on GIS and one-user one-file data according to claim 1, wherein the step S5 specifically comprises: Performing feature space mapping on the historical cases in the semantic rule knowledge base based on the plot comprehensive evaluation index to construct an initial retrieval set of candidate suggestions; Performing similarity calculation on the current parcel feature vector and the triplet parcel features, service actions and implementation effects in the semantic rule knowledge base by using a cosine similarity algorithm to screen out historical cases with similarity higher than a preset threshold; carrying out semantic clustering on service actions corresponding to the screened historical cases, and merging similar actions by adopting a K-means algorithm to generate a structured candidate suggestion category set; Performing a path matching algorithm based on the knowledge graph on the candidate suggestion category set to identify an optimal decision path which is most matched with the current land parcel features; And generating a candidate suggestion set comprising business actions such as volume rate increasing application, compensation standard adjustment, placement scheme optimization and the like according to the output result of the optimal decision path, and adding a corresponding implementation effect predicted value for each suggestion.

Description

Method for analyzing, studying and judging value of removed land block based on GIS and one-user one-file data Technical Field The invention relates to a data intelligent analysis and decision support technology in the field of urban updating, in particular to a method for analyzing and judging the value of a removed land block based on GIS and one-user one-file data. Background In the field of urban updating and demolition land development value evaluation, a main stream technical scheme generally adopts a fusion analysis method based on GIS space data and one-user one-file multi-source attribute data, and development potential and implementation cost of a to-be-demolished land are exponentially presented by constructing a quantitative evaluation model. The outstanding problems in the prior art include that a semantic gap exists between a model output result and a business decision, so that an evaluation result is difficult to be directly adopted by a business department to influence decision conversion efficiency, and in addition, in the aspect of model interpretation, a systematic presentation of evaluation result support logic and an implementation path is lacking, and the transparency and the credibility of a decision process are influenced due to the lack of a correlation causal network between evaluation data and actual business actions. Disclosure of Invention The application provides a method for analyzing and judging the value of a removed land block based on GIS and one-user one-file data, which aims to solve one of the problems or the problems of the prior art mentioned in the background art. The application provides a method for analyzing and judging the value of a removed land block based on GIS and one-user one-file data, which specifically comprises the following steps: S1, collecting attribute data of 'one household one file' and GIS space data in an urban updating project, wherein the attribute data comprises property types, population structures and history compensation records, the space data comprises land block boundaries, volume rate and peripheral matching information, and the attribute data of one household one file is subjected to standardized cleaning and the GIS space data is subjected to coordinate system one processing. And S2, based on a GIS space analysis technology, establishing a mapping relation between one-user one-file data and GIS space data elements through a space connection tool, and converting unstructured archive data into a structured analysis unit with geographic coordinate association by adopting a containing and intersecting topological rule. And S3, constructing a multidimensional migration plot evaluation model, dividing a development value index calculation module and an implementation cost calculation module, and generating a plot comprehensive evaluation index by adopting a weighted summation method, wherein the weight parameters of the development value module are dynamically adjusted according to the regional characteristics. And S4, performing text mining on the historical city updating project file by using a natural language processing technology, extracting decision rules of compensation standards, volume rate adjustment and placement schemes, and constructing a semantic rule knowledge base containing triples of land block features, business actions and implementation effects. S5, constructing a semantic reasoning engine comprising a feature mapping module, a rule matching module and a logic reasoning module, inputting the plot comprehensive evaluation index into the semantic reasoning engine, calculating the similarity between the current plot feature and the historical plot feature in the semantic rule knowledge base to search an optimal decision path, and reasoning by the logic reasoning module based on the mapping relation between the business action and the implementation effect in the triplet to generate a candidate suggestion set comprising volume rate improvement application, compensation standard adjustment and placement scheme optimization. S6, based on the block space location, the weight complexity and the peripheral matched maturity characteristics, multidimensional decision tree analysis is performed on the candidate suggestion set, and development strategy combination with implementation priority ranking and a corresponding policy adaptation scheme are output. And S7, generating a visual decision map, displaying a causal association network of 'evaluation index-plot characteristic-recommendation strategy', marking supporting evidence sources of each recommendation strategy, and automatically generating an evaluation report abstract containing value basis and implementation path. And S8, establishing a manual feedback closed-loop mechanism, recording deviation reasons and updating decision weight parameters in a semantic rule knowledge base when the actual decision deviates from a system recommendation scheme, and optimizing the case ma