CN-120124725-B - Multi-source geographic space big data knowledge graph construction method and system for intelligent city planning

CN120124725BCN 120124725 BCN120124725 BCN 120124725BCN-120124725-B

Abstract

The invention discloses a multi-source geographic space big data knowledge graph construction method and system for intelligent city planning, wherein the method comprises the steps of collecting original data of a vector layer, a grid layer, a text layer and an image layer, and preprocessing to respectively obtain vector data, grid data, text data and image data; and extracting respective entity sets, relationship sets, attribute sets and alignment modes in a vector layer, a grid layer, a text layer and an image layer in the target city by utilizing a DeepLabV3+ semantic segmentation model according to the set and alignment modes constructed in the knowledge graph body layer, and forming a multi-source geographic space big data knowledge graph by utilizing the NLP model and the GIS platform based on the software platform. The high-efficiency integration and accurate alignment of the heterogeneous data are realized through the unified reference entity, and the defects of the traditional method in data integration efficiency, entity alignment precision and intelligent analysis capability are overcome.

Inventors

DUAN YUXI
CHEN BIYU
MA YAOHONG

Assignees

武汉大学

Dates

Publication Date: 20260512
Application Date: 20250224

Claims (7)

1. A multi-source geographic space big data knowledge graph construction method for intelligent city planning is characterized by comprising the following steps: The method comprises the steps of collecting original data of a vector layer, a grid layer, a text layer and an image layer, and preprocessing the original data to respectively obtain vector data, grid data, text data and image data, wherein the vector data at least comprises road data and administrative area data in a research area; constructing a knowledge graph ontology layer, wherein the construction of the knowledge graph ontology layer comprises constructing a vector entity set, a vector attribute set and a vector relation set according to vector data, constructing a grid entity set, a grid attribute set and a grid relation set according to grid data, constructing a text entity set, a text attribute set and a text relation set according to text data, constructing an image entity set, an image attribute set and an image relation set according to image data, aligning the vector entity set, the grid entity set, the text entity set and the image entity set by taking a street name as a reference entity to obtain an entity set, and determining an attribute set and a relation set according to the position relation of entity elements and entity elements in the obtained entity set and the attribute of the entity elements; the image entity set, the image attribute set and the image relation set are constructed by marking geographic entities in image data based on the ID of sampling points along the road in a research area to form the image entity set, extracting the green vision rate, the building vision rate and the vehicle vision rate in the image data by using a deep learning model DeepLabV & lt+ & gt to form the image attribute set; According to a set and alignment mode constructed in a knowledge graph body layer, a DeepLabV & lt+ & gt semantic segmentation model is utilized, a natural language processing NLP technical model and a GIS platform are based on a software platform, respective entity sets, relationship sets, attribute sets and alignment modes in a vector layer, a grid layer, a text layer and an image layer in a target city are extracted, a multi-source geographic space big data knowledge graph is formed, the GIS platform is used as a space analysis tool for extracting entities, attributes and relationships of the vector layer and extracting entities, attributes and relationships of the grid layer, and the natural language processing NLP technical model is used for extracting the entities, attributes and relationships of the text layer.
2. The method for constructing the multi-source geospatial big data knowledge graph for intelligent city planning of claim 1 wherein constructing the set of vector entities, the set of vector attributes, and the set of vector relationships comprises: extracting a road name, an OSM (open object model) identifier and an administrative division from the vector data to form a vector entity set; Calculating the topological attribute and the geometric attribute of each entity element in the vector entity set to form a vector attribute set, wherein the topological attribute at least comprises adjacent roads and road lengths, and the geometric attribute at least comprises longitude and latitude; And establishing the space and attribute relation of the category, the length, the nearest road, the Euclidean distance and the administrative division of each entity element in the vector entity set to form a vector relation set.
3. The method for constructing the multi-source geospatial big data knowledge graph for intelligent city planning of claim 1 wherein constructing the set of grid entities, the set of grid attributes, and the set of grid relationships comprises: Generating sampling point IDs along the road in the research area and combining time fields to form a grid entity set; extracting grid attribute values of all grid entity elements in a grid entity set in remote sensing data in a research area to form a grid attribute set; geographic positions of all grid entity elements in a grid entity set in remote sensing data in a research area are extracted, and geographic position relations of all grid entity elements in the grid entity set are established to form a grid relation set.
4. The method for constructing a multi-source geospatial big data knowledge graph for intelligent city planning in accordance with claim 1, wherein the text data comprises store criticizing data that can be separated into latitude and longitude by at least natural language processing techniques.
5. The method for constructing a multi-source geospatial big data knowledge graph for intelligent city planning in accordance with claim 4, wherein constructing the set of text entities, the set of text relationships, and the set of text attributes comprises: extracting shop names in a research area from the text data to form a text entity set; Extracting text attributes of all text entity elements in the text entity set from the text data to form a text attribute set A t ; and constructing geographical relations among shops, nearest roads, administrative regions where shops are located and longitude and latitude according to the names of shops in the text data to form a text relation set.
6. The method for constructing a multi-source geospatial big data knowledge graph for intelligent city planning in accordance with claim 1, wherein said image data comprises picture data of green view rate, building view rate and vehicle view rate extracted from a study area.
7. A multi-source geospatial big data knowledge graph construction system for intelligent city planning, comprising: The system comprises a data collection processing module, a data processing module and a platform, wherein the data collection processing module is used for collecting the original data of a vector layer, a grid layer, a text layer and an image layer and preprocessing the original data to respectively obtain vector data, grid data, text data and image data, wherein the vector data at least comprises road data and administrative area data in a research area; the ontology layer construction module is used for constructing a vector entity set, a vector attribute set and a vector relation set according to vector data, constructing a grid entity set, a grid attribute set and a grid relation set according to raster data, constructing a text entity set, a text attribute set and a text relation set according to text data, constructing an image entity set, an image attribute set and an image relation set according to image data, aligning the vector entity set, the grid entity set, the text entity set and the image entity set by taking a street name as a reference entity to obtain an entity set, determining an attribute set and a relation set according to the obtained position relation of entity elements and attributes of the entity elements in the entity set, and constructing a knowledge map ontology layer; The knowledge graph construction module is used for utilizing DeepLabV3+ semantic segmentation models according to a set and an alignment mode constructed in a knowledge graph body layer, extracting respective entity sets, relationship sets, attribute sets and the alignment mode in a vector layer, a grid layer, a text layer and an image layer in a target city by using a natural language processing NLP technical model and a GIS platform based on a software platform to form a multi-source geographic space big data knowledge graph, wherein the GIS platform is used as a space analysis tool for extracting entities, attributes and relationships of the vector layer and extracting the entities, attributes and relationships of the grid layer, and the natural language processing NLP technical model is used for extracting the entities, the attributes and the relationships of the text layer.

Description

Multi-source geographic space big data knowledge graph construction method and system for intelligent city planning Technical Field The invention belongs to the technical field of knowledge maps and smart cities, and particularly relates to a multi-source geographic space big data knowledge map construction method and system for smart city planning. Background With the rapid advancement of the global urbanization process, the construction of smart cities puts higher demands on urban planning. However, conventional city planning tools and methods, such as Geographic Information Systems (GIS) and remote sensing technology, face significant challenges for integration of multi-source heterogeneous geographic data. These data sources (such as remote sensing images, sensor networks and socioeconomic data) have significant differences in terms of spatial resolution, timeliness and data type, making them difficult to apply effectively in smart city planning where real-time and accuracy requirements are high. In addition, the current multi-source geographic data processing still has the problems of low efficiency, high complexity and the like, which limits the actual effectiveness of the multi-source geographic data processing in dynamic decision making and urban planning to a certain extent. Therefore, how to realize efficient integration, accurate analysis and intelligent application of multi-source heterogeneous geographic data becomes a technical bottleneck to be solved in smart city planning. Aiming at the problems, the geographic knowledge graph (GeoKG) is used as an emerging technical means, and the deep integration and intelligent analysis of the data are realized by converting the multi-source geographic information into a structured knowledge form. GeoKG can effectively integrate vector data, raster data, text data, image data and other data types, accurately describe geographic entities and relations thereof by utilizing a semantic modeling technology, and provide a brand-new data support framework for smart city planning. However, currently GeoKG still faces many challenges in practical applications, particularly in terms of data accuracy, entity alignment, and multi-modal data seamless interfacing. How to overcome these problems by efficient modeling and analysis methods is a key direction to advance intelligent development of smart city planning. Disclosure of Invention The invention aims at solving the problems in the prior art, and provides a multi-source geographic space big data knowledge graph construction method and system for intelligent city planning. According to an aspect of the present invention, there is provided a multi-source geospatial big data knowledge graph construction method for intelligent city planning, including: Raw Data of a vector layer, a grid layer, a text layer and an image layer are collected and preprocessed to respectively obtain vector Data V, grid Data R, text Data T and image Data I; Constructing a knowledge graph ontology layer, wherein the construction of the knowledge graph ontology layer comprises constructing a vector entity set E v according to vector Data V, Vector attribute set A v and vector relation set R v, and grid entity set E r is constructed from grid Data R, The grid attribute set A r and the grid relation set R r construct a text entity set E t from text Data T, A text attribute set A t and a text relation set R t, and an image entity set E i is constructed from image Data I, Image attribute set A i and image relation set R i, vector entity set E v, grid entity set E r, and street name as reference entity, aligning the text entity set E t and the image entity set E i to obtain an entity set E= { E v,Er,Et,Ei }, and determining an attribute set A and a relationship set R according to the obtained entity elements in the entity set E and the position relationship and the attribute of the entity elements; And extracting respective entity sets, relationship sets, attribute sets and alignment modes in a vector layer, a grid layer, a text layer and an image layer in the target city by utilizing a DeepLabV3+ semantic segmentation model, an NLP model and a GIS platform based on a software platform according to the set and alignment modes constructed in the knowledge graph body layer to form a multi-source geographic space big data knowledge graph. According to the technical scheme, the high-efficiency integration and accurate alignment of heterogeneous data are achieved through the unified reference entity, the defects of the traditional method in data integration efficiency, entity alignment precision and intelligent analysis capability are overcome, automation and high efficiency of data processing are guaranteed through the combination of DeepLabV3+ semantic segmentation model, natural language processing technology and GIS space analysis tool, a brand new data management and analysis framework is provided for smart city planning by utilizing the