A method and system for automatic analysis of a business district for commercial real estate planning

Abstract

The application relates to the technical field of data analysis and intelligent planning, and discloses a business circle automatic analysis method and system for commercial real estate planning, wherein the method comprises the following steps: acquiring mobile phone signaling data, map trajectory data, commercial point of interest (POI) data and traffic transfer node data; performing a business circle space-time knowledge graph construction task; performing a candidate business circle identification preprocessing task; performing a business circle hierarchical correction task; performing business circle overlap degree analysis; and outputting a commercial real estate planning analysis result. Compared with the prior art which mainly relies on administrative division to delimit a business circle, especially under the condition of mixed distribution of commercial formats, the technical problem that real business circle boundary identification cannot be realized is solved. According to the application, the automatic identification of a core business circle, a secondary business circle and a potential development business circle in a target region is realized by constructing a business circle space-time knowledge graph and combining business circle hierarchical correction, and the accuracy of commercial real estate planning analysis is improved.

Description

Technical Field

[0001] This invention relates to the field of data analysis and intelligent planning technology, and in particular to an automatic business district analysis method and system for commercial real estate planning. Background Technology

[0002] Currently, the methods for analyzing business districts in commercial real estate planning still have significant shortcomings. For example, existing technologies typically rely on administrative boundaries, fixed service radii, manual experience-based delineation, single heat maps, or static POI density distributions to identify and assess business districts. While these methods can reflect the level of commercial activity in a region to some extent, they are essentially static and coarse-grained analytical approaches, failing to accurately describe the dynamic changes in business district boundaries in real urban environments. Especially in urban spaces where multiple transportation modes such as subways, buses, pedestrian networks, and shared mobility overlap, passenger arrival paths, dwell behaviors, migration directions, and cross-regional flow relationships exhibit significant spatiotemporal fluctuations and network coupling. Traditional analytical methods often fail to effectively reflect the true business district structure driven by "traffic organization—passenger migration—commercial agglomeration."

[0003] Furthermore, existing technologies for business district analysis mostly focus on a single data source, such as using only mobile signaling data to characterize customer traffic intensity, only using POI data to characterize commercial supply density, or only using map trajectory data to analyze travel routes. They lack the ability to uniformly model the relationships between mobile signaling data, map trajectory data, POI data, and transportation transfer node data. Since different data sources reflect different aspects of customer dwell time, customer migration, commercial layout, and traffic flow capacity, the inability to uniformly organize and collaboratively analyze this heterogeneous data can easily lead to problems such as significant deviations in business district boundary identification, unclear business district hierarchy, difficulty in quantifying competitive relationships between business districts, and inaccurate commercial positioning results for target sites.

[0004] Therefore, there is an urgent need for an automatic business district analysis method for commercial real estate planning that can still automatically identify the real business district boundaries driven by multiple sources of customer flow, automatically determine the business district level, and jointly analyze the overlap, competition index, and radiation capacity of business districts, even in situations where urban transportation transfer relationships are complex, customer flow sources are diverse, business formats are mixed and distributed, and business district boundaries are dynamically changing. This method would improve the accuracy, dynamic adaptability, and decision support capabilities of commercial real estate planning and analysis. Summary of the Invention

[0005] To address the aforementioned technical shortcomings, the purpose of this invention is to propose an automatic business district analysis method for commercial real estate planning. This method aims to solve the technical problem that existing technologies mainly rely on administrative divisions to delineate business districts, especially when business formats are mixed, making it impossible to identify the true boundaries of business districts.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides an automatic business district analysis method for commercial real estate planning.

[0007] The automated business district analysis method for commercial real estate planning includes: Step S10: Obtain mobile phone signaling data, map trajectory data, commercial point of interest (POI) data and transportation transfer node data corresponding to the target city area, and perform the spatiotemporal observation sequence construction task in combination with the multi-source passenger flow unified rasterization construction mechanism to output passenger flow intensity characteristics, business structure characteristics and transportation accessibility characteristics. Step S20: Based on passenger flow intensity characteristics, business format structure characteristics, and traffic accessibility characteristics, a multi-source entity relationship mapping mechanism is used to perform the task of constructing a spatiotemporal knowledge graph of the business district, and the graph embedding feature set and spatial association weight matrix are output. Step S30: Based on the graph embedding feature set and spatial association weight matrix, a spatiotemporal clustering fusion mechanism is used to perform the candidate business district identification preprocessing task, and output the initial business district area set, boundary unit set, central unit set and migration connection sub-graph; Step S40: Based on the initial set of business district regions, the set of boundary units, the set of central units, and the migration connection subgraph, perform the business district classification correction task and output the set of business district boundaries and the set of business district level labels; Step S50: Based on the set of business district boundaries and the set of business district level labels, conduct business district overlap analysis, competition index analysis, and radiation capacity analysis, and output the commercial real estate planning analysis results.

[0008] Preferably, step S10, which involves acquiring mobile phone signaling data, map trajectory data, points of interest (POI) data, and transportation transfer node data corresponding to the target city area, and combining this with a multi-source passenger flow unified rasterization construction mechanism to perform a spatiotemporal observation sequence construction task, and outputting passenger flow intensity characteristics, business structure characteristics, and transportation accessibility characteristics, specifically includes: Step S101: Obtain mobile phone signaling data, map trajectory data, commercial point of interest (POI) data, and transportation transfer node data corresponding to the target city area. Perform anonymization, outlier removal, timestamp alignment, and coordinate transformation on the mobile phone signaling data, map trajectory data, POI data, and transportation transfer node data to obtain a standardized spatiotemporal raw dataset. Step S102: Divide the target city area into multiple spatial analysis units, map the standardized spatiotemporal raw dataset to each spatial analysis unit, and statistically analyze the rasterized observation parameters in each spatial analysis unit within the pre-approved spatial analysis time window. The rasterized observation parameters include the number of people staying, the number of people arriving, the number of people leaving, the average length of stay, the number of transfers, and the number of commercial points of interest. Step S103: Based on the rasterized observation parameters, the normalized weighted statistical method is used to calculate the passenger flow intensity characteristics, business format structure characteristics, and traffic accessibility characteristics of each spatial analysis unit. Among them, the passenger flow intensity characteristics are used to characterize the ability and activity level of each spatial analysis unit to attract consumer traffic within the corresponding time window; the business format structure characteristics are used to characterize the type composition, functional mixing degree, and commercial supply characteristics of commercial interest points within each spatial analysis unit; and the traffic accessibility characteristics are used to characterize the accessibility efficiency between each spatial analysis unit and transportation transfer nodes and the ability to attract external passenger flow.

[0009] Preferably, step S20, which involves constructing a spatiotemporal knowledge graph of a business district based on passenger flow intensity characteristics, business format structure characteristics, and traffic accessibility characteristics, and using a multi-source entity relationship mapping mechanism, and outputting the graph embedding feature set and spatial association weight matrix, specifically includes: Step S201: Based on passenger flow intensity characteristics, business structure characteristics, and traffic accessibility characteristics, a heterogeneous entity extraction method is used to construct a multi-type entity set, which includes spatial analysis unit entities, commercial interest point entities, transportation transfer node entities, and time slice entities; based on the multi-type entity set, semantic association modeling is performed using relation triples to output the initial business district spatiotemporal knowledge graph. Step S202: Extract passenger flow migration relationships separately from map trajectory data, and establish passenger flow migration edges for the initial business district spatiotemporal knowledge graph based on the passenger flow migration relationships using the weighted graph edge construction method, thereby obtaining the spatial association weight matrix; Step S203: The initial business district spatiotemporal knowledge graph and spatial association weight matrix are updated by using a graph embedding learning mechanism to obtain a graph embedding feature set. Finally, the graph embedding feature set and spatial association weight matrix are output.

[0010] Preferably, in step S20, the initial business district spatiotemporal knowledge graph includes at least the following entity relationships: the inclusion relationship between spatial analysis unit entities and commercial interest point entities; the adjacency and reachability relationship between spatial analysis unit entities and transportation transfer node entities; the spatial adjacency relationship between adjacent spatial analysis unit entities; the temporal evolution relationship between the same spatial analysis unit entities under different time slices; and the passenger flow migration relationship between different spatial analysis unit entities.

[0011] Preferably, in step S30, the step of performing the candidate business district identification preprocessing task based on the spatiotemporal clustering fusion mechanism using the map embedding feature set and spatial correlation weight matrix, and outputting the initial business district region set, boundary unit set, central unit set, and migration connection subgraph, specifically includes: Step S301: Based on the graph embedding feature set, the NumPy and NetworkX libraries in Python are used to calculate and process the row graph structure features, outputting the business district agglomeration potential energy parameter set for each spatial analysis unit. The business district agglomeration potential energy parameter set includes passenger flow intensity potential energy parameter, dwell time potential energy parameter, connectivity centrality parameter, and graph embedding local density. Among them, the connectivity centrality parameter is used to characterize the degree of connectivity of each spatial analysis unit in the overall passenger flow migration network; the graph embedding local density is used to characterize the degree of local agglomeration of each spatial analysis unit in the knowledge graph feature space. Step S302: Based on the set of agglomeration potential energy parameters of the business district, select candidate core units from each spatial analysis unit, and perform clustering expansion processing on the candidate core units using a constrained spatiotemporal density clustering method based on the spatial correlation weight matrix, and output candidate clusters; when the number of units in the candidate clusters exceeds the preset retention threshold, retain the candidate core units as the initial set of business district areas; Step S303: For the initial business district set, use the shapely library of Python to extract its boundary cell set and central cell set, and construct the migration connection subgraph based on the boundary cell set and central cell set using the migration connection subgraph construction method. Finally, output the initial business district set, boundary cell set, central cell set and migration connection subgraph.

[0012] Preferably, step S40, which involves performing a business district classification correction task based on the initial business district area set, boundary unit set, center unit set, and migration connection subgraph, and outputting the business district boundary set and business district level label set, specifically includes: Step S401: Calculate the core score, secondary support score and growth score of each initial business district in the initial business district area set. The core score is used to represent the dominant ability of the corresponding business district area to attract customers and commercial resources in the area. The secondary support score is used to represent the ability of the corresponding business district area to take over the core business district. The growth score is used to represent the expansion potential of the corresponding business district area within the preset statistical period. Step S402: Preset core score threshold conditions, secondary support score threshold conditions, and growth score threshold conditions; perform business district level determination based on the core score, secondary support score, and growth score of each initial business district area; mark the initial business district areas that meet the core score threshold conditions as core business districts; mark the initial business district areas that meet the secondary support score threshold conditions as secondary business districts; mark the initial business district areas that meet the growth score threshold conditions as potential development business districts; and output a set of business district level labels. Step S403: Obtain the first migration connection strength between central units based on the central unit set and the migration connection subgraph; perform business district center connection and merging processing on central units with the first migration connection strength higher than the preset merging threshold; obtain the second migration connection strength between boundary units based on the boundary unit set and the migration connection subgraph; perform boundary clipping and removal processing on boundary units with the second migration connection strength lower than the preset removal threshold; finally output the business district boundary set.

[0013] Preferably, step S50, which involves performing business district overlap analysis, competition index analysis, and radiation capacity analysis based on the business district boundary set and business district level label set, and outputting the commercial real estate planning analysis results, specifically includes: Step S501: Based on the set of business district boundaries, a spatial overlap detection method is used to detect the boundary intersection relationship. The boundary intersection relationship and customer group cross relationship between any two business districts are output. Based on the boundary intersection relationship and customer group cross relationship, a business district overlap index is constructed using a weighted normalization method. The business district overlap index is used to characterize the degree of overlap between two business districts in terms of spatial coverage and customer flow source. Step S502: Construct a competition index based on the set of business district level labels and the business district overlap index using a piecewise weighted function mapping method. The competition index is used to characterize the intensity of competition among different business districts around the target consumer group, commercial supply resources, and location attractiveness. Construct a radiation capacity index based on the set of business district level labels and the business district overlap index using a Sigmoid function hierarchical mapping method. The radiation capacity index is used to characterize the ability of the corresponding business district to attract and diffuse surrounding spatial analysis units, external customer flow sources, and target plots. Step S503: Input the business district overlap index, competition index and radiation capacity index into the pre-set commercial site selection decision model, and the commercial site selection decision model outputs the commercial real estate planning analysis results.

[0014] This invention also provides an automatic business district analysis system for commercial real estate planning, comprising: The multi-source spatiotemporal data construction module is used to acquire mobile phone signaling data, map trajectory data, commercial point of interest (POI) data and transportation transfer node data corresponding to the target city area. Combined with the multi-source passenger flow unified rasterization construction mechanism, it performs the spatiotemporal observation sequence construction task and outputs passenger flow intensity characteristics, business structure characteristics and transportation accessibility characteristics. The spatiotemporal knowledge graph construction module is used to perform the spatiotemporal knowledge graph construction task of the business district based on passenger flow intensity characteristics, business structure characteristics and traffic accessibility characteristics, and adopts a multi-source entity relationship mapping mechanism to output the graph embedding feature set and spatial association weight matrix. The candidate business district identification module is used to perform the candidate business district identification preprocessing task based on the spatiotemporal clustering fusion mechanism using the graph embedding feature set and spatial correlation weight matrix, and outputs the initial business district area set, boundary unit set, central unit set and migration connection subgraph; The business district classification correction module is used to perform business district classification correction tasks based on the initial business district area set, boundary unit set, center unit set, and migration connection subgraph, and output the business district boundary set and business district level label set. The planning analysis output module is used to perform business district overlap analysis, competition index analysis, and radiation capacity analysis based on the business district boundary set and business district level label set, and output the commercial real estate planning analysis results.

[0015] The present invention also provides an automatic business district analysis device for commercial real estate planning, comprising: a memory, a processor, and an automatic business district analysis program for commercial real estate planning stored in the memory and executable on the processor. When the automatic business district analysis program for commercial real estate planning is executed by the processor, it implements an automatic business district analysis method for commercial real estate planning.

[0016] The present invention also provides a computer program product, including an automatic business district analysis program for commercial real estate planning, wherein the automatic business district analysis program for commercial real estate planning implements the automatic business district analysis method for commercial real estate planning when executed by a processor.

[0017] The beneficial effects of this invention are as follows: By uniformly rasterizing mobile phone signaling data, map trajectory data, commercial point of interest (POI) data, and transportation transfer node data, and further constructing a spatiotemporal knowledge graph of business districts, a graph embedding feature set, and a spatial correlation weight matrix, this invention can break through the traditional static analysis method of dividing business districts based on administrative divisions, fixed radii, or single heat values. In scenarios with complex transportation transfer relationships, intertwined passenger flow migration paths, and obvious mixed distribution of business formats, this invention can more accurately identify the true boundaries of business districts and achieve automatic differentiation between core business districts, secondary business districts, and potential development business districts, thereby significantly improving the consistency between business district identification results and actual consumption activity distribution.

[0018] Based on the identification of business district boundaries, this invention further combines business district overlap analysis, competition index analysis, and radiation capacity analysis to simultaneously output quantitative analysis results for commercial real estate planning. It can not only reflect the spatial overlap and competition intensity between different business districts, but also characterize the attraction and diffusion capacity of each business district to the target plot and surrounding areas. This transforms the decision-making process for commercial real estate site selection, functional positioning, and business format configuration from experience-based judgment to data-driven decision-making, thereby improving the scientific nature, pertinence, and reliability of commercial real estate planning. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the first embodiment of an automatic business district analysis method for commercial real estate planning according to the present invention.

[0021] Figure 2 This is a schematic diagram of an equipment for an automatic business district analysis method for commercial real estate planning according to the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Example 1: As Figure 1 The diagram shown is a flowchart of the first embodiment of the automatic business district analysis method for commercial real estate planning according to the present invention, which presents the first embodiment of the automatic business district analysis method for commercial real estate planning according to the present invention.

[0024] In the first embodiment, the automatic business district analysis method for commercial real estate planning includes: Step S10: Obtain mobile phone signaling data, map trajectory data, commercial point of interest (POI) data and transportation transfer node data corresponding to the target city area, and perform the spatiotemporal observation sequence construction task in combination with the multi-source passenger flow unified rasterization construction mechanism to output passenger flow intensity characteristics, business structure characteristics and transportation accessibility characteristics. It should be noted that this step involves first mapping multi-source data with different origins, granularities, and representations onto a unified spatiotemporal analysis framework, and then extracting fundamental features that truly reflect the conditions for the formation of business districts within this unified framework. Specifically, mobile phone signaling data is mainly used to reflect changes in the number of people staying at different locations within the target city area at different time periods, the strength of their staying behavior, and the activity level of the population; map trajectory data is mainly used to reflect the arrival paths, departure paths, migration directions, and cross-regional flow relationships of people; commercial point of interest (POI) data is mainly used to reflect the spatial layout of commercial facilities, the composition of business types, and the density of commercial supply; and transportation transfer node data is mainly used to reflect the supporting role of transportation facilities such as subway stations, bus hubs, and pedestrian connections in attracting passenger flow. Since the above different data collectively describe the formation process of business districts from four dimensions—"personal stay," "personal migration," "commercial supply," and "transportation attraction"—this step integrates the above data into a unified spatiotemporal observation sequence through a multi-source passenger flow unified rasterization construction mechanism, providing a consistent, comparable, and computable basic input for subsequent knowledge graph construction and candidate business district identification.

[0025] Understandably, the multi-source passenger flow unified grid construction mechanism refers to first dividing the target city area into multiple spatial analysis units based on its area, commercial density, and subsequent calculation accuracy requirements. Then, using a preset analysis time window as the statistical period, different data sources are uniformly collected, aligned, and parameters extracted within each spatial analysis unit. The spatial analysis unit can be a regular square grid, a hexagonal grid, or a local unit adapted to road network segmentation; the analysis time window can be set to 15 minutes, 30 minutes, 1 hour, or divided according to weekday peak, weekday off-peak, and weekend peak periods. Through this processing method, raw data from different sources are all organized into a basic statistical structure of "the same spatial unit + the same time window," so that the subsequently output passenger flow intensity characteristics, business structure characteristics, and traffic accessibility characteristics are no longer isolated indicators, but rather composite characteristic results with a unified spatiotemporal reference system. In other words, the three types of features output in this step essentially correspond to the commercial customer attraction capacity, commercial supply organization capacity, and traffic diversion support capacity of each spatial analysis unit within the preset time window. These three factors together determine whether the spatial analysis unit has the basic conditions to form the core or periphery of a business district.

[0026] It should be understood that "passenger flow intensity characteristics" can be comprehensively formed by observational parameters such as the number of people staying, the number of people arriving, the number of people leaving, the average length of stay, and the frequency of migration per unit time. Its purpose is to characterize the degree of passenger flow aggregation and the level of passenger flow activity within a specific spatial analysis unit during the statistical period. "Business format structure characteristics" can be comprehensively formed by parameters such as the number of commercial interest points, the density of commercial interest points, the proportion of catering, retail, entertainment, and office services, etc., and its purpose is to characterize the commercial function types and business format mixing status of the spatial analysis unit. "Transportation accessibility characteristics" can be comprehensively formed by parameters such as the road network distance from the spatial analysis unit to the nearest transportation transfer node, the number of transfer lines, the average transfer time, the convenience of walking connections, and the frequency of external passenger flow inflow, and its purpose is to characterize the efficiency of the spatial analysis unit in receiving external consumer passenger flow. Furthermore, the anonymization, outlier removal, timestamp alignment, and coordinate transformation of the original data in this step are to avoid feature distortion caused by inconsistencies in sampling accuracy, timestamp format, and spatial coordinate benchmarks between different data sources, thereby ensuring the engineering usability and statistical stability of the output features. Compared to traditional methods that rely solely on heat maps or POI density maps for business district analysis, this step emphasizes first converting multi-source heterogeneous data into a unified spatiotemporal observation sequence, and then extracting basic features based on this sequence. This approach more accurately reflects the dynamic mechanisms of real business district formation. For example, taking a core area of ​​a city as an example, suppose the target city area is divided into spatial analysis units with sides of 200 meters, and a 30-minute analysis time window is used. Within a single spatial analysis unit, during weekday evening peak hours, the statistics show 680 people staying, 310 arriving, 255 departing, an average stay duration of 42 minutes, and 128 transfers. This unit contains 86 points of interest (POIs), with restaurants accounting for 0.34%, retail for 0.29%, entertainment for 0.17%, and lifestyle services for 0.20%. The nearest subway transfer point is 180 meters away, with an average walking time of 3.2 minutes. Based on the above parameters, it can be determined that this spatial analysis unit has high passenger flow intensity, strong business mix characteristics, and good transportation accessibility characteristics, thus making it more likely to be identified as a candidate core unit in subsequent steps. Conversely, if another spatial analysis unit has a large number of commercial points of interest, but low number of people staying, short average stay time, and a long distance from transportation transfer nodes, then although it has a certain commercial supply capacity, it may not have the conditions to form a core business district, and is more likely to be identified as a secondary unit or peripheral unit in subsequent processing.

[0027] Step S20: Based on passenger flow intensity characteristics, business format structure characteristics, and traffic accessibility characteristics, a multi-source entity relationship mapping mechanism is used to perform the task of constructing a spatiotemporal knowledge graph of the business district, and the graph embedding feature set and spatial association weight matrix are output. It should be noted that this step maps these features to various entities that can express urban commercial spatial relationships and their interrelationships, thereby constructing a spatiotemporal knowledge graph of business districts with spatiotemporal semantic association capabilities. In other words, the passenger flow intensity features, business format structure features, and transportation accessibility features output in step S10 are no longer merely static values ​​in this step, but are assigned to spatial analysis unit entities, commercial interest point entities, transportation transfer node entities, and time slice entities, respectively. This gives each entity not only its own attributes but also structured relationships with other entities. In this way, the originally scattered features of "passenger flow intensity," "business format distribution," and "transportation accessibility" can be integrated into a graph structure model that reflects the formation mechanism of business districts, thus providing intermediate results for subsequent candidate business district identification that can be used for cluster relationship inference and association strength measurement.

[0028] Understandably, the multi-source entity relationship mapping mechanism refers to first abstracting the analysis objects in the target urban area into entities based on the three types of features output in step S10, and then modeling the relationships between different entities. Specifically, spatial analysis unit entities are used to carry the comprehensive passenger flow and commercial characteristics within a certain area grid or a certain spatial unit; commercial interest point entities are used to carry the category, quantity, distribution density, and business type attributes of the corresponding commercial facilities; transportation transfer node entities are used to carry the accessibility and flow guidance attributes of facilities such as subway stations, bus transfer stations, and integrated transportation hubs; and time slice entities are used to carry the dynamic change characteristics of each spatial analysis unit at different time periods. After the entities are constructed, relationships such as "spatial analysis unit entity - containment - commercial interest point entity", "spatial analysis unit entity - adjacent accessibility - transportation transfer node entity", "spatial analysis unit entity - spatial adjacency - spatial analysis unit entity", "spatial analysis unit entity - temporal evolution - time slice entity", and "spatial analysis unit entity - passenger flow migration - spatial analysis unit entity" are established through relational triples. In this way, the three types of features extracted in the previous step are converted into node attributes and edge attributes in the graph structure, so that the knowledge graph can not only describe "what features a certain unit has", but also further describe "which units this unit has migration connections with, which transportation nodes it has reachability relationships with, and which commercial facilities it has supply connections with".

[0029] For example, taking a large commercial core area in a city as an example, assume that step S10 has obtained the passenger flow intensity characteristics, business structure characteristics, and transportation accessibility characteristics of several adjacent spatial analysis units. Specifically, spatial analysis unit 1 has a high number of residents and average dwell time during the evening peak hours, with its internal commercial interest points mainly consisting of catering, retail, and entertainment, and it is relatively close to a subway transfer node. Spatial analysis unit 2 is adjacent to spatial analysis unit 1; although its passenger flow intensity is slightly lower, it has a higher proportion of office service-related commercial interest points, and there is a strong map trajectory migration relationship between it and spatial analysis unit 1. Spatial analysis unit 3 also has a certain density of commercial interest points, but it lacks a significant passenger flow migration relationship with the previous two, and it is far from transportation transfer nodes. In this step, spatial analysis units 1, 2, and 3 are respectively constructed as spatial analysis unit entities, and relationships are established with their corresponding commercial interest point entities, transportation transfer node entities, and time slice entities. Subsequently, based on the smaller road network distance, higher passenger flow migration intensity, and stronger functional complementarity between the first and second spatial analysis units, a larger spatial association weight can be obtained. Conversely, the third spatial analysis unit, due to its weaker migration connection and poorer traffic support with the first and second units, has a lower corresponding association value in the spatial association weight matrix. After further graph embedding learning, the vector distance between the first and second spatial analysis units in the graph embedding feature space becomes closer, while the vector position of the third spatial analysis unit is relatively offset. Therefore, the graph embedding feature set and spatial association weight matrix output in this step can provide direct evidence for candidate business district identification in the subsequent step S30, making it easier to identify the first and second spatial analysis units as the same candidate business district cluster, while excluding the third spatial analysis unit from this cluster.

[0030] Step S30: Based on the graph embedding feature set and spatial association weight matrix, a spatiotemporal clustering fusion mechanism is used to perform the candidate business district identification preprocessing task, and output the initial business district area set, boundary unit set, central unit set and migration connection sub-graph; It should be noted that this step further transforms the map embedding feature set and spatial association weight matrix output from step S20 into regional-level results that can be used for business district division. That is, it moves from "what kind of association exists between units" to "which spatial analysis units should collectively constitute the same candidate business district." In other words, step S20 focuses more on expressing the semantic relationships and the strength of connections between spatial analysis units, while this step, based on this expression, performs cluster discrimination, cluster structure expansion, and preliminary boundary extraction on the spatial analysis units, thereby outputting an initial set of business district regions, a set of boundary units, a set of central units, and a migration connection subgraph. That is to say, this step does not directly provide the final business district boundary, but rather completes the pre-identification of candidate business district regions and the pre-stratification of the internal structure of the regions, providing structured input for the subsequent business district classification correction and boundary refinement in step S40.

[0031] Understandably, the spatiotemporal clustering fusion mechanism refers to jointly calculating the semantic proximity reflected by the graph embedding feature set and the spatial connectivity, migration intensity, and functional similarity reflected by the spatial association weight matrix to form a comprehensive evaluation of the clustering tendency of each spatial analysis unit. Specifically, the graph embedding feature set can reflect the vector positional relationship of each spatial analysis unit in the overall knowledge graph. If two spatial analysis units are close in the graph embedding space, it indicates that they have a high degree of comprehensive similarity in terms of passenger flow activity patterns, business supply relationships, and transportation support conditions. The spatial association weight matrix further reflects the degree of road network connectivity, passenger flow migration intensity, and transportation transfer connectivity of the two units in the real urban space. By fusing these two types of information, the misjudgment of "geographically adjacent but functionally fragmented" when dividing business districts solely based on spatial adjacency can be avoided, as can the distortion problem of "similar features but spatial discontinuity" when dividing business districts solely based on semantic similarity can also be avoided. Therefore, the initial set of business districts output in this step retains both spatial continuity and consistency in commercial functions and customer flow organization.

[0032] It should be understood that the initial business district set output in this step refers to a candidate business cluster area composed of multiple adjacent or strongly connected spatial analysis units; the boundary unit set refers to a set of spatial analysis units located on the periphery of the candidate business cluster area, in direct contact with the external area, and which need to be judged for retention or removal in subsequent boundary refinement; the central unit set refers to a set of key spatial analysis units with high customer flow intensity, strong connectivity centrality, or high local density of map embedding in the corresponding initial business district area; and the migration connection subgraph refers to a local migration relationship network constructed around the interior and boundary of a certain initial business district area, used to describe the strength of customer flow migration connections between central units and boundary units, and between boundary units. Furthermore, the "set of potential energy parameters for business district agglomeration" extracted in this step is essentially a quantitative expression of the role of each spatial analysis unit in the formation of candidate business districts. Among these, the customer flow intensity potential energy parameter characterizes the spatial analysis unit's ability to attract consumers; the dwell time potential energy parameter characterizes the spatial analysis unit's capacity to support consumer activity; the connectivity centrality parameter characterizes the spatial analysis unit's hub status in the overall migration network; and the local density of the graph embedding characterizes the clustering density of the spatial analysis unit in the comprehensive feature space. Only through the combined effect of these four types of potential energy parameters can we more accurately identify which spatial analysis units are suitable as candidate core units, which are only suitable as candidate boundary units, and which should be excluded from the initial business district area.

[0033] For example, taking the area surrounding a commercial complex in a city as an example, assume that step S20 has already output the set of graph embedding features and spatial association weight matrix of multiple adjacent spatial analysis units. Among them, spatial analysis unit A is located near the main entrance of a large shopping mall, with high potential energy parameters of passenger flow intensity, long average dwell time, high connectivity centrality, and high local density of graph embedding; spatial analysis unit B is located in the block north of spatial analysis unit A. Although the passenger flow intensity is slightly lower, the spatial association weight between it and spatial analysis unit A is large, and the graph embedding vector distance is relatively close; spatial analysis unit C is located in the surrounding office building area. Its business structure is somewhat complementary to that of spatial analysis unit A, but the passenger flow migration intensity between it and spatial analysis unit A is weak; spatial analysis unit D is geographically adjacent, but its accessibility is poor, its dwell time is short, and its local density of graph embedding is low. In this step, spatial analysis unit A is first identified as a candidate core unit. Spatial analysis unit B, due to its strong spatial correlation and close map embedding location with spatial analysis unit A, is usually expanded and incorporated into the same initial business district area. Spatial analysis unit C may be temporarily excluded from the initial business district area due to insufficient migration connections or retained only as a weakly correlated edge unit. Spatial analysis unit D is more likely to be directly filtered. Subsequently, after the initial business district area is formed, spatial analysis unit A is identified as the central unit, and some units in spatial analysis unit B near the outer roads may be identified as boundary units. At the same time, a migration connection subgraph is constructed based on the passenger flow migration relationships between spatial analysis units A and B, as well as between boundary units.

[0034] Step S40: Based on the initial set of business district regions, the set of boundary units, the set of central units, and the migration connection subgraph, perform the business district classification correction task and output the set of business district boundaries and the set of business district level labels; It should be noted that this step involves further discrimination, screening, and correction of the initial business district area set output in step S30, transforming the candidate business district areas from "pre-identified results" into "formal business district results usable for commercial real estate planning analysis." Specifically, although step S30 has already output the initial business district area set, boundary unit set, central unit set, and migration connection sub-graph, this result is essentially still in a candidate state. It may contain areas with significant differences in business district intensity, areas with excessively wide or narrow boundaries, and weakly correlated areas temporarily included due to short-term fluctuations in customer flow. Therefore, this step needs to further complete two key tasks based on the aforementioned output: firstly, to determine the level of each initial business district area to distinguish between core business districts, secondary business districts, and potential development business districts; secondly, to correct the initial business district boundaries based on the migration connection strength between central units and boundary units, the aggregation potential of the boundary units themselves, and regional continuity. Through this process, the final output set of business district boundaries is no longer a rough outline of candidate clusters, but a formal boundary result that more accurately reflects the scope of commercial space; the output set of business district level labels is no longer just a simple cluster number, but a hierarchical identification result that can directly serve commercial real estate site selection, positioning and comparative analysis.

[0035] It should be understood that the task of classifying and revising business districts refers to simultaneously addressing two aspects: "regional value assessment" and "boundary structure revision," assuming that candidate business districts have already been formed. First, regarding regional value assessment, a core strength score, a secondary support score, and a growth potential score need to be calculated for each initial business district. The core strength score primarily reflects the district's dominant position within the overall commercial network, generally determined by the customer flow intensity, average dwell time, business density, and the strength of primary migration connections between central units within the district. The secondary support score primarily reflects the district's ability to collaboratively absorb adjacent strong business districts, generally determined by the degree of boundary connectivity between the district and surrounding core business districts, the stability of connections between boundary units and central units, and the degree of business supplementation. The growth potential score primarily reflects the district's potential to expand into a higher-level business district within a preset statistical period, generally determined by the trend of customer flow growth, the trend of improved traffic flow, and the activity level of newly added commercial interest points. Second, regarding boundary structure revision, based on the set of central units and the set of boundary units, the outer edge units of the initial business district need to be retained, eliminated, or incorporated. In other words, if a boundary unit is included in the candidate business district during the initial clustering, but its migration connection with the central unit is weak and its own agglomeration potential is low, it should be eliminated in this step; conversely, if an external neighboring unit is not directly included in the initial business district area in step S30, but it has a strong migration connection with the central unit and has strong customer flow and commercial support characteristics, it can be incorporated into the formal business district boundary in this step. For example, taking three candidate areas in a city center as an example, assuming that step S30 has already output the initial business district area A, the initial business district area B, and the initial business district area C, the initial business district area A contains multiple units surrounding large shopping malls. The units in the central unit set have high customer flow intensity and long average dwell time, and the first migration connection strength between the central units is relatively high. The initial business district area B is located on the periphery of the initial business district area A. Its internal commercial interest points are mainly catering, community retail, and supporting services. There is a relatively stable migration connection between its boundary units and the central units of the initial business district area A. The initial business district area C is currently small in scale and has a limited number of central units, but it is adjacent to a newly built subway transfer node, and the customer flow growth rate and the proportion of newly added commercial interest points have increased significantly in recent statistical periods. In this step, initial business district A can be marked as a core business district because its core score is significantly higher than the preset threshold. Although initial business district B has a lower core score than initial business district A, it can be marked as a secondary business district because of its higher secondary support score. Although initial business district C currently has a small customer flow, it has a high growth potential score and can therefore be marked as a potential development business district.Meanwhile, for certain boundary units surrounding the initial business district A, if their second migration connection strength with the central unit is low, and the boundary unit itself has a short dwell time and insufficient business support in multiple time windows, these boundary units will be eliminated in this step. However, external units with a high first migration connection strength with the central unit and spatially continuous adjacency can be merged into the business district boundary. Ultimately, after the above-mentioned level determination and boundary correction processing, the output set of business district boundaries can more accurately reflect the actual scope of different levels of business districts in the target city area, and the output set of business district level labels can more directly reflect the functional positioning and comparative value of different business districts in commercial real estate planning.

[0036] Step S50: Based on the set of business district boundaries and the set of business district level labels, conduct business district overlap analysis, competition index analysis, and radiation capacity analysis, and output the commercial real estate planning analysis results.

[0037] It should be noted that this step further transforms the set of business district boundaries and the set of business district level labels output in step S40 into analytical results that can directly serve commercial real estate planning decisions. In other words, the previous steps have already completed the identification, boundary correction, and hierarchical labeling of actual business districts. This step, however, builds upon this foundation to further answer three types of questions more relevant to planning applications: "Do different business districts overlap?", "How strong is the competitive relationship between different business districts?", and "What is the level of attraction and diffusion capacity of each business district to the target plot and surrounding areas?". Compared to simply outputting static identification results indicating which business district a certain area belongs to, this step emphasizes the quantitative expression of the relationships between business districts and the deduction of the planning value of the target plot. Therefore, it is a key implementation link in this invention, moving from "business district identification" to "planning decision-making." Through this step, it no longer merely provides the spatial outline of business districts but can further output explanatory conclusions for commercial real estate site selection, business format configuration, and functional positioning, making the aforementioned identification results truly valuable for business applications.

[0038] Understandably, the business district overlap analysis refers to calculating the degree of overlap between two business districts in terms of spatial coverage and customer flow origins, based on the spatial intersection and customer group crossover relationships between their boundaries. If the boundaries of two business districts have a large overlapping area geographically, and the customer sources, dwell behaviors, and migration paths in the corresponding overlapping area also have high similarity, it indicates a strong overlap in their target consumer group coverage, and the business district overlap is high. Conversely, if two business districts are geographically adjacent but have limited boundary overlap and significant differences in customer sources, their business district overlap is low. Furthermore, the so-called competition index analysis refers to quantifying the competitive relationship between two business districts around similar consumer demands and commercial resources, based on the business district overlap analysis and combined with the business district level label set, business format similarity, location proximity, and customer flow migration relationships. Generally speaking, when two core business districts have a high degree of overlap, similar business formats, and significant customer flow separation, their competition index will be relatively high. Conversely, when a core business district and a secondary business district have a more dominant and subordinate relationship, their competition index will be relatively low. As for radiation capacity analysis, it refers to quantifying each business district's ability to attract external customers, stimulate surrounding consumption activities, and influence the commercial value of target sites based on the set of business district level labels, the set of business district boundaries, and the strength of migration connections between the business district and surrounding units. The stronger the radiation capacity, the more significant the business district's commercial driving effect on the surrounding area, and the higher the value of the commercial environment to which the target site is located.

[0039] Example 2: Furthermore, the present invention provides an automatic business district analysis system for commercial real estate planning, employing an automatic business district analysis method for commercial real estate planning as described in the above embodiments, which can solve a technical problem related to automatic business district analysis for commercial real estate planning. The beneficial effects of the automatic business district analysis system for commercial real estate planning provided by the present invention are the same as those of the automatic business district analysis method for commercial real estate planning provided in the above embodiments, and other technical features of the automatic business district analysis system for commercial real estate planning are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0040] Example 3: This invention provides an automatic business district analysis device for commercial real estate planning. Please refer to... Figure 2An automatic business district analysis device for commercial real estate planning includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to perform an automatic business district analysis method for commercial real estate planning as described in Embodiment 1 above. The automatic business district analysis device for commercial real estate planning in this embodiment may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and in-vehicle terminals (e.g., in-vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. This automatic business district analysis device for commercial real estate planning is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this invention. An automated business district analysis device for commercial real estate planning may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to a program stored in a read-only memory 1002 or a program loaded from a storage device 1003 into a random access memory 1004. The random access memory 1004 also stores various programs and data required for the operation of the automated business district analysis device for commercial real estate planning. The processing unit 1001, the read-only memory 1002, and the random access memory 1004 are interconnected via a bus 1005. An I / O interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows an automated commercial district analysis device for commercial real estate planning to communicate wirelessly or wiredly with other devices to exchange data. While the figure shows an automated commercial district analysis device for commercial real estate planning with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0041] Example 4: This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method for automatic business district analysis in commercial real estate planning. The computer program product provided by this invention can solve a technical problem related to automatic business district analysis in commercial real estate planning. Compared with the prior art, the beneficial effects of the computer program product provided by this invention are the same as those of the automatic business district analysis method for commercial real estate planning provided in the above embodiments, and will not be repeated here.

[0042] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this invention.

[0043] It should be understood that the various parts disclosed in this invention can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0044] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. An automatic business district analysis method for commercial real estate planning, characterized in that, The methods include: Step S10: Obtain mobile phone signaling data, map trajectory data, commercial point of interest (POI) data and transportation transfer node data corresponding to the target city area, and perform the spatiotemporal observation sequence construction task in combination with the multi-source passenger flow unified rasterization construction mechanism to output passenger flow intensity characteristics, business structure characteristics and transportation accessibility characteristics. Step S20: Based on passenger flow intensity characteristics, business format structure characteristics, and traffic accessibility characteristics, a multi-source entity relationship mapping mechanism is used to perform the task of constructing a spatiotemporal knowledge graph of the business district, and the graph embedding feature set and spatial association weight matrix are output. Step S30: Based on the graph embedding feature set and spatial association weight matrix, a spatiotemporal clustering fusion mechanism is used to perform the candidate business district identification preprocessing task, and output the initial business district area set, boundary unit set, central unit set and migration connection sub-graph; Step S40: Based on the initial set of business district regions, the set of boundary units, the set of central units, and the migration connection subgraph, perform the business district classification correction task and output the set of business district boundaries and the set of business district level labels; Step S50: Based on the set of business district boundaries and the set of business district level labels, conduct business district overlap analysis, competition index analysis, and radiation capacity analysis, and output the commercial real estate planning analysis results.

2. The automatic business district analysis method for commercial real estate planning as described in claim 1, characterized in that, Step S10 involves acquiring mobile phone signaling data, map trajectory data, points of interest (POI) data, and transportation transfer node data corresponding to the target city area. It also involves constructing a spatiotemporal observation sequence using a multi-source passenger flow unified rasterization mechanism, outputting passenger flow intensity characteristics, business structure characteristics, and transportation accessibility characteristics. Specifically, this includes: Step S101: Obtain mobile phone signaling data, map trajectory data, commercial point of interest (POI) data, and transportation transfer node data corresponding to the target city area. Perform anonymization, outlier removal, timestamp alignment, and coordinate transformation on the mobile phone signaling data, map trajectory data, POI data, and transportation transfer node data to obtain a standardized spatiotemporal raw dataset. Step S102: Divide the target city area into multiple spatial analysis units, map the standardized spatiotemporal raw dataset to each spatial analysis unit, and statistically analyze the rasterized observation parameters in each spatial analysis unit within the pre-approved spatial analysis time window. The rasterized observation parameters include the number of people staying, the number of people arriving, the number of people leaving, the average length of stay, the number of transfers, and the number of commercial points of interest. Step S103: Based on the rasterized observation parameters, the normalized weighted statistical method is used to calculate the passenger flow intensity characteristics, business format structure characteristics, and traffic accessibility characteristics of each spatial analysis unit. Among them, the passenger flow intensity characteristics are used to characterize the ability and activity level of each spatial analysis unit to attract consumer traffic within the corresponding time window; the business format structure characteristics are used to characterize the type composition, functional mixing degree, and commercial supply characteristics of commercial interest points within each spatial analysis unit; and the traffic accessibility characteristics are used to characterize the accessibility efficiency between each spatial analysis unit and transportation transfer nodes and the ability to attract external passenger flow.

3. The automatic business district analysis method for commercial real estate planning as described in claim 1, characterized in that, In step S20, based on passenger flow intensity characteristics, business format structure characteristics, and traffic accessibility characteristics, a multi-source entity relationship mapping mechanism is used to perform the task of constructing a spatiotemporal knowledge graph of the business district, and the step of outputting the graph embedding feature set and spatial association weight matrix specifically includes: Step S201: Based on passenger flow intensity characteristics, business structure characteristics, and traffic accessibility characteristics, a heterogeneous entity extraction method is used to construct a multi-type entity set, which includes spatial analysis unit entities, commercial interest point entities, transportation transfer node entities, and time slice entities; based on the multi-type entity set, semantic association modeling is performed using relation triples to output the initial business district spatiotemporal knowledge graph. Step S202: Extract passenger flow migration relationships separately from map trajectory data, and establish passenger flow migration edges for the initial business district spatiotemporal knowledge graph based on the passenger flow migration relationships using the weighted graph edge construction method, thereby obtaining the spatial association weight matrix; Step S203: The initial business district spatiotemporal knowledge graph and spatial association weight matrix are updated by using a graph embedding learning mechanism to obtain a graph embedding feature set. Finally, the graph embedding feature set and spatial association weight matrix are output.

4. The automatic business district analysis method for commercial real estate planning as described in claim 3, characterized in that, In step S20, the initial spatiotemporal knowledge graph of the business district includes at least the following entity relationships: the inclusion relationship between spatial analysis unit entities and commercial interest point entities; the adjacency and reachability relationship between spatial analysis unit entities and transportation transfer node entities; and the spatial adjacency relationship between adjacent spatial analysis unit entities. The temporal evolution relationship between entities in the same spatial analysis unit under different time slices; the passenger flow migration relationship between entities in different spatial analysis units.

5. The automatic business district analysis method for commercial real estate planning as described in claim 2, characterized in that, In step S30, the preprocessing task for candidate business district identification is performed using a spatiotemporal clustering fusion mechanism based on the map embedding feature set and spatial correlation weight matrix, outputting the initial business district region set, boundary unit set, central unit set, and migration connection subgraph. Specifically, this includes: Step S301: Based on the graph embedding feature set, the NumPy and NetworkX libraries in Python are used to calculate and process the row graph structure features, outputting the business district agglomeration potential energy parameter set for each spatial analysis unit. The business district agglomeration potential energy parameter set includes passenger flow intensity potential energy parameter, dwell time potential energy parameter, connectivity centrality parameter, and graph embedding local density. Among them, the connectivity centrality parameter is used to characterize the degree of connectivity of each spatial analysis unit in the overall passenger flow migration network; the graph embedding local density is used to characterize the degree of local agglomeration of each spatial analysis unit in the knowledge graph feature space. Step S302: Based on the set of agglomeration potential energy parameters of the business district, select candidate core units from each spatial analysis unit, and perform clustering expansion processing on the candidate core units using a constrained spatiotemporal density clustering method based on the spatial correlation weight matrix, and output candidate clusters; when the number of units in the candidate clusters exceeds the preset retention threshold, retain the candidate core units as the initial set of business district areas; Step S303: For the initial business district set, use the shapely library of Python to extract its boundary cell set and central cell set, and construct the migration connection subgraph based on the boundary cell set and central cell set using the migration connection subgraph construction method. Finally, output the initial business district set, boundary cell set, central cell set and migration connection subgraph.

6. The automatic business district analysis method for commercial real estate planning as described in claim 1, characterized in that, Step S40, which involves performing a business district classification correction task based on the initial business district area set, boundary unit set, center unit set, and migration connection subgraph, and outputting the business district boundary set and business district level label set, specifically includes: Step S401: Calculate the core score, secondary support score and growth score of each initial business district in the initial business district area set. The core score is used to represent the dominant ability of the corresponding business district area to attract customers and commercial resources in the area. The secondary support score is used to represent the ability of the corresponding business district area to take over the core business district. The growth score is used to represent the expansion potential of the corresponding business district area within the preset statistical period. Step S402: Preset core score threshold conditions, secondary support score threshold conditions, and growth score threshold conditions; perform business district level determination based on the core score, secondary support score, and growth score of each initial business district area; mark the initial business district areas that meet the core score threshold conditions as core business districts; mark the initial business district areas that meet the secondary support score threshold conditions as secondary business districts; mark the initial business district areas that meet the growth score threshold conditions as potential development business districts; and output a set of business district level labels. Step S403: Obtain the first migration connection strength between central units based on the central unit set and the migration connection subgraph; perform business district center connection and merging processing on central units with the first migration connection strength higher than the preset merging threshold; obtain the second migration connection strength between boundary units based on the boundary unit set and the migration connection subgraph; perform boundary clipping and removal processing on boundary units with the second migration connection strength lower than the preset removal threshold; finally output the business district boundary set.

7. The automatic business district analysis method for commercial real estate planning as described in claim 1, characterized in that, Step S50, which involves performing business district overlap analysis, competition index analysis, and radiation capacity analysis based on the business district boundary set and business district level label set, and outputting the commercial real estate planning analysis results, specifically includes: Step S501: Based on the set of business district boundaries, a spatial overlap detection method is used to detect the boundary intersection relationship. The boundary intersection relationship and customer group cross relationship between any two business districts are output. Based on the boundary intersection relationship and customer group cross relationship, a business district overlap index is constructed using a weighted normalization method. The business district overlap index is used to characterize the degree of overlap between two business districts in terms of spatial coverage and customer flow source. Step S502: Construct a competition index based on the set of business district level labels and the business district overlap index using a piecewise weighted function mapping method. The competition index is used to characterize the intensity of competition among different business districts around the target consumer group, commercial supply resources, and location attractiveness. Construct a radiation capacity index based on the set of business district level labels and the business district overlap index using a Sigmoid function hierarchical mapping method. The radiation capacity index is used to characterize the ability of the corresponding business district to attract and diffuse surrounding spatial analysis units, external customer flow sources, and target plots. Step S503: Input the business district overlap index, competition index and radiation capacity index into the pre-set commercial site selection decision model, and the commercial site selection decision model outputs the commercial real estate planning analysis results.

8. An automatic business district analysis system for commercial real estate planning, applied to the automatic business district analysis method for commercial real estate planning as described in any one of claims 1 to 7, characterized in that, The automated business district analysis system for commercial real estate planning includes: The multi-source spatiotemporal data construction module is used to acquire mobile phone signaling data, map trajectory data, commercial point of interest (POI) data and transportation transfer node data corresponding to the target city area. Combined with the multi-source passenger flow unified rasterization construction mechanism, it performs the spatiotemporal observation sequence construction task and outputs passenger flow intensity characteristics, business structure characteristics and transportation accessibility characteristics. The spatiotemporal knowledge graph construction module is used to perform the spatiotemporal knowledge graph construction task of the business district based on passenger flow intensity characteristics, business structure characteristics and traffic accessibility characteristics, and adopts a multi-source entity relationship mapping mechanism to output the graph embedding feature set and spatial association weight matrix. The candidate business district identification module is used to perform the candidate business district identification preprocessing task based on the spatiotemporal clustering fusion mechanism using the graph embedding feature set and spatial correlation weight matrix, and outputs the initial business district area set, boundary unit set, central unit set and migration connection subgraph; The business district classification correction module is used to perform business district classification correction tasks based on the initial business district area set, boundary unit set, center unit set, and migration connection subgraph, and output the business district boundary set and business district level label set. The planning analysis output module is used to perform business district overlap analysis, competition index analysis, and radiation capacity analysis based on the business district boundary set and business district level label set, and output the commercial real estate planning analysis results.

9. An automatic business district analysis device for commercial real estate planning, characterized in that, The automatic business district analysis device for commercial real estate planning includes: a memory, a processor, and an automatic business district analysis program for commercial real estate planning stored in the memory and executable on the processor. When the automatic business district analysis program for commercial real estate planning is executed by the processor, it implements the automatic business district analysis method for commercial real estate planning according to any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes an automatic business district analysis program for commercial real estate planning, which, when executed by a processor, implements an automatic business district analysis method for commercial real estate planning as described in any one of claims 1 to 7.

Images