A virtual park construction and enterprise matching method, medium and device

Abstract

This invention relates to the field of data processing technology, and in particular to a method, medium, and device for constructing virtual parks and matching enterprises. It establishes associations between unmatched enterprises and buildings based on spatial distance and calculates matching confidence, providing high-quality spatial units and reliable basic weights for virtual park construction. Virtual parks are generated by spatially clustering candidate building nodes, and the overall confidence of the park is calculated by integrating the overall confidence of the buildings, the number of enterprises associated with each building, and the standard deviation of spatial distance between buildings. This ensures that the generation of virtual parks considers both the positive contributions of data source credibility and enterprise clustering scale, while also introducing a penalty mechanism for spatial compactness, thus improving the quality of virtual park generation. Finally, the association confidence is obtained by combining the matching confidence and the overall confidence of the park, allowing the enterprise attribution results to comprehensively reflect the credibility of the enterprise-building match and the spatial clustering effectiveness of the virtual park itself, thereby improving the reliability of enterprise spatial attribution.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, medium, and equipment for constructing and matching virtual parks with enterprises. Background Technology

[0002] In industrial park management and enterprise spatial attribution analysis, it is often necessary to spatially correlate scattered enterprise data with physical parks. However, in real-world scenarios, a large number of enterprises cannot be matched with any physical park, such as those located in scattered office buildings, industrial incubators, co-working spaces, or enterprise clusters that have not yet been formally designated as parks. For these enterprises, existing technologies mainly rely on manual experience to define virtual park boundaries or use simple clustering methods based on fixed spatial distance thresholds for aggregation.

[0003] However, manual delineation is highly subjective and inefficient, making it difficult to adapt to large-scale, dynamically changing enterprise spatial distribution scenarios. Furthermore, traditional clustering methods based on fixed distance thresholds only consider spatial proximity, ignoring the enterprise size information associated with the building data itself and the degree of spatial density between buildings. This results in the generated virtual parks lacking reliable quantification, and the accuracy of the association results between enterprises and virtual parks is low, making it difficult to meet the needs of refined spatial governance.

[0004] Therefore, how to build a highly credible virtual park and connect it with enterprises has become an urgent technical problem to be solved. Summary of the Invention

[0005] To address the aforementioned technical problems, the present invention provides a method for constructing a virtual park and matching enterprises, which includes the following steps: S1: Obtain the set of enterprises that have not been matched with physical parks, the building dataset from several data sources, and the physical park data. Each unmatched enterprise in the enterprise set has the coordinates of its center point. The building dataset includes several buildings, each with the coordinates of its center point and a building's overall confidence score. The physical park data includes information on the buildings covered by each physical park.

[0006] S2, based on the spatial distance between the center point coordinates of each enterprise and the center point coordinates of each building, associate each unmatched enterprise with the building with the closest spatial distance, and calculate the matching confidence between the unmatched enterprise and the associated building based on the spatial distance.

[0007] S3 selects buildings from the building dataset that are associated with at least one unmatched enterprise and are not covered by a physical park as candidate building nodes.

[0008] S4. Cluster the candidate building nodes according to the coordinates of the building center point to generate at least one building cluster, and determine each building cluster as a virtual park.

[0009] S5. For each virtual park, the overall confidence level of the virtual park is calculated based on the overall confidence level of each building in the virtual park, the number of unmatched enterprises associated with each building, and the standard deviation of the spatial distance between any two buildings in the virtual park. The standard deviation of the spatial distance is used to characterize the density of the spatial distribution between buildings.

[0010] S6 assigns each unmatched enterprise in the enterprise set to the corresponding virtual park through the associated building, and obtains the association confidence between each unmatched enterprise and the associated virtual park based on the matching confidence and the overall park confidence.

[0011] The present invention also provides a non-transitory computer-readable storage medium storing at least one instruction or at least one program, wherein the at least one instruction or at least one program is loaded and executed by a processor to implement the above-described virtual park construction and enterprise matching method.

[0012] The present invention also provides an electronic device, including a processor and the aforementioned non-transitory computer-readable storage medium.

[0013] This invention has at least the following beneficial effects: It establishes the association between unmatched enterprises and buildings based on the spatial distance between the enterprise's central point and the building's central point, and calculates the matching confidence level, providing high-quality, weighted spatial units and reliable basic weights for virtual park construction; it generates virtual parks by spatially clustering candidate building nodes, and calculates the park's comprehensive confidence level by integrating the building's overall confidence level, the number of enterprises associated with the building, and the standard deviation of the spatial distance between buildings. This ensures that the generation of virtual parks considers both the credibility of the data source and the positive contribution of the enterprise clustering scale, while also introducing a spatial compactness penalty mechanism, achieving a multi-dimensional quantitative evaluation of virtual park quality and avoiding the problems of loose clustering and artificially high confidence levels caused by relying solely on spatial distance; by combining the matching confidence level and the park's comprehensive confidence level to obtain the association confidence level between unmatched enterprises and virtual parks, the enterprise attribution results comprehensively reflect the credibility of the matching between unmatched enterprises and buildings, as well as the spatial clustering effectiveness of the virtual park itself, improving the accuracy and reliability of enterprise spatial attribution. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying 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.

[0015] Figure 1 This is a flowchart of a virtual park construction and enterprise matching method provided in Embodiment 1 of the present invention. Detailed Implementation

[0016] 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.

[0017] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It is understood that, where appropriate, the terms used to distinguish similar objects can be interchanged so that the invention can also be implemented in other embodiments besides the illustrated or described embodiments. Furthermore, the terms "including," "having," and any variations are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products, or devices.

[0018] Example 1 This first embodiment provides a method for constructing a virtual campus and matching enterprises, such as... Figure 1 As shown, the method for constructing a virtual campus and matching it with enterprises includes the following steps: S1 retrieves the set of enterprises that are not matched to physical parks, building datasets from several data sources, and physical park data.

[0019] Among them, physical parks are existing formal industrial parks, science parks, development zones, etc., with clear spatial boundaries and jurisdiction. Unmatched enterprises are enterprises that have not established a spatial affiliation with existing physical parks, such as those located in scattered office buildings, industrial incubators, co-working spaces, enterprise clusters that have not been formally included in parks, or scattered enterprises in rural areas; each unmatched enterprise in the enterprise set has corresponding coordinates of the enterprise's center point, such as latitude and longitude.

[0020] Specifically, to distinguish between unmatched enterprises and those matched to physical parks, this embodiment obtains the coordinates of the enterprise's center point for each initial enterprise and the coverage area of ​​the boundary polygon of each physical park. For each initial enterprise, it determines whether its enterprise center point coordinates fall inside the boundary polygon, or whether the distance from the enterprise center point coordinates to the centroid of the boundary polygon is within a preset proximity threshold. If the above conditions are met, the initial enterprise is associated with the corresponding physical park; otherwise, the initial enterprise not associated with any physical park is considered an unmatched enterprise, and virtual park construction and secondary association matching are then performed. The specific value of the preset proximity threshold can be set by the implementer according to the actual situation. For example, in this embodiment, the preset proximity threshold is 20 meters.

[0021] Buildings are physical structures in a city that can be used for business offices or industrial activities, including office buildings, commercial buildings, towers, and office buildings. The building dataset includes several buildings, each with its own center point coordinates and overall building confidence score. This overall building confidence score is quantified based on the authority of the data source through a probability complementary cumulative multiplication fusion formula, reflecting the credibility of the building data.

[0022] The physical park data includes building information covered by each physical park, which is used to exclude buildings that have already been covered, and to avoid spatial overlap and jurisdictional conflicts between virtual parks and physical parks.

[0023] In one specific implementation, the data source includes official registration data, map service provider POI data, and OSM open-source building outline data. S1 includes the following steps: Obtain the initial confidence level for each data source.

[0024] For any given building, the overall building confidence level is calculated using a probability complementary cumulative multiplication fusion formula based on several data sources corresponding to the building and the initial confidence level of each data source.

[0025] Different data sources possess varying levels of authority and accuracy. In this embodiment, these differences are represented by an initial confidence level, which reflects the variations during multi-source fusion, allowing highly authoritative data sources to contribute more to the final confidence level. For instance, officially registered data has the highest authority, with an initial confidence level set at 0.9, making it the most reliable source of building information. Map service provider POI data has broad coverage and is updated frequently, with an initial confidence level set at 0.7. OSM open-source building outline data may contain errors or delays, with an initial confidence level set at 0.6.

[0026] When the same building is recorded by multiple data sources, each data source provides independent and complementary evidence. This embodiment uses a probabilistic complementary cumulative multiplication fusion formula to calculate the building's overall confidence level based on several data sources corresponding to the building and the initial confidence level of each data source. When multiple data sources are consistent, the overall confidence level is higher than any single source. However, when multiple data sources conflict, the overall confidence level will be limited due to the low confidence level of a certain source. This provides a reliable basic weight for subsequent virtual park confidence level calculations.

[0027] Specifically, the building's overall confidence level C buliding =1-∏ n i=1 (1-C i ), where C buliding C represents the overall confidence level of the building. i Let be the initial confidence level of the i-th data source corresponding to the building, n be the total number of data sources corresponding to the building, and ∏ be the product function.

[0028] The above-mentioned multi-source data fusion and confidence quantification provide high-quality, weighted spatial units and reliable basic weights for the construction of virtual parks.

[0029] S2, based on the spatial distance between the center point coordinates of each enterprise and the center point coordinates of each building, associate each unmatched enterprise with the building with the closest spatial distance, and calculate the matching confidence between the unmatched enterprise and the associated building based on the spatial distance.

[0030] This method calculates the spherical straight-line distance between the coordinates of the enterprise's center point and the building's center point. This distance is used to quantify the spatial proximity between the enterprise and the building. The closer the spatial distance, the higher the probability that the unmatched enterprise belongs to that building. Therefore, each unmatched enterprise is associated with the building with the closest spatial distance. In this embodiment, the Havesing formula is used to calculate this spatial distance, and the unit is meters.

[0031] Match confidence represents the credibility of spatial matching between unmatched enterprises and associated buildings. It is used to quantify the reliability of the association between unmatched enterprises and buildings. The value ranges from 0 to 1, and match confidence is negatively correlated with spatial distance.

[0032] In one specific embodiment, S2 includes the following steps: Based on the nearest spatial distance and the preset baseline attenuation distance, the matching confidence between the unmatched enterprise and the associated building is calculated using an exponential decay function.

[0033] Among them, the spatial distance between unmatched enterprises and buildings and the confidence of matching show a monotonically decreasing relationship: the closer the distance, the higher the confidence; the farther the distance, the lower the confidence. Moreover, exponential decay can quickly reduce the confidence of long-distance matching.

[0034] In this embodiment, the exponential decay function is: C match =e^(-d / D0), where C match The matching confidence level between the unmatched enterprise and the associated building is given by d, where d is the spatial distance between the unmatched enterprise and the associated building, and D0 is the preset baseline attenuation distance.

[0035] The specific value of the preset reference attenuation distance can be set by the implementer according to the actual situation. For example, in this embodiment, the preset reference attenuation distance is 150 meters.

[0036] In one specific embodiment, S2 further includes the following steps: If the matching confidence level is less than the preset confidence threshold, the association between the unmatched enterprise and the associated building is deemed invalid, and the unmatched enterprise is removed from the enterprise set.

[0037] In cases where the distance between an unmatched enterprise and its associated building is too far and the matching confidence is lower than the preset confidence threshold, forcibly assigning it to a building would introduce a lot of noise and reduce the association quality. Therefore, it is considered that the spatial relationship between the unmatched enterprise and its associated building is insufficient to support a reliable association. Thus, the association between the unmatched enterprise and its associated building is deemed invalid, and a deletion operation is triggered to remove the unmatched enterprise from the enterprise set to save computing resources. The unmatched enterprise will no longer participate in the subsequent construction of the virtual park and will be defined as a standalone scattered enterprise or a rural scattered enterprise.

[0038] The specific value of the preset reliability threshold can be set by the implementer according to the actual situation. For example, in this embodiment, the preset reliability threshold is 0.16.

[0039] The above-mentioned mechanisms, including nearest neighbor assignment, exponential decay confidence, and threshold filtering, achieve quality control of the association between unmatched enterprises and buildings, providing clean and reliable association information for subsequent candidate building screening and virtual park confidence calculation.

[0040] S3 selects buildings from the building dataset that are associated with at least one unmatched enterprise and are not covered by a physical park as candidate building nodes.

[0041] The above-mentioned dual-condition screening significantly reduced the size of the candidate building node set, lowered the computational complexity of subsequent clustering steps, and improved the business relevance and practical application value of virtual park construction.

[0042] S4. Cluster the candidate building nodes according to the coordinates of the building center point to generate at least one building cluster, and determine each building cluster as a virtual park.

[0043] Buildings that are spatially close to each other often form shared service areas, share supporting facilities, and attract similar types of businesses, thus they can be considered a functional whole. Clustering algorithms can automatically handle clusters of different shapes and densities without requiring the number of clusters to be specified in advance.

[0044] Therefore, in this embodiment, for spatial point data of building types, a density-based clustering algorithm, such as DBSCAN, is used based on the coordinates of the building's center point. By setting the neighborhood radius and the minimum number of core points, the spherical distance between any two candidate building nodes is calculated. Clustering is performed based on the neighborhood radius and the minimum number of core points, dividing buildings that meet the density-connected condition into the same cluster, generating at least one building cluster. Each building cluster is defined as a virtual park, and a unique identifier is generated for it.

[0045] The specific values ​​of the neighborhood radius and the minimum number of core points can be set by the implementer according to the actual situation. For example, in this embodiment, the neighborhood radius is set to 200 meters and the minimum number of core points is 2.

[0046] The above-mentioned unsupervised clustering of candidate building nodes is performed based on the spatial location and density clustering algorithm of buildings. Each building cluster is corresponding to a virtual park. Subsequent confidence calculation and enterprise affiliation are performed on the virtual park as the unit, forming a virtual park management system parallel to the physical park, which improves the integrity of the association between enterprises and parks.

[0047] S5. For each virtual park, the overall confidence level of the virtual park is calculated based on the overall confidence level of each building in the virtual park, the number of unmatched enterprises associated with each building, and the standard deviation of the spatial distance between any two buildings in the virtual park. The standard deviation of the spatial distance is used to characterize the density of the spatial distribution between buildings.

[0048] In one specific implementation, the overall confidence level of the park is positively correlated with the overall confidence level of the buildings and the number of unmatched enterprises, and negatively correlated with the standard deviation of spatial distance.

[0049] Specifically, for each virtual park, the spatial distance between all pairs of buildings within that virtual park is calculated, and the standard deviation of these spatial distances is also calculated to obtain the spatial distance standard deviation, which quantifies the dispersion of building distribution. Correspondingly, the smaller the spatial distance standard deviation, the denser the building distribution, and the larger the spatial distance standard deviation, the looser the building distribution.

[0050] The credibility of a virtual park is determined by the quality and clustering of its buildings. Specifically, the higher the data credibility of the buildings themselves (overall building confidence level) and the higher the actual number of enterprises occupying the buildings (the number of associated unmatched enterprises), the higher the overall confidence level of the virtual park. The looser the spatial distribution between buildings (the larger the standard deviation of spatial distance), the weaker the clustering effect of the virtual park, the worse its legitimacy as a park, and the lower its overall confidence level should be.

[0051] In this embodiment, C virtual =e -σ ×Σ(N b ×C b ) / ΣN b , where C virtual C represents the overall confidence level of the virtual park. b Let N be the overall building confidence level for each building within the virtual campus. b Let σ represent the number of unmatched companies associated with each building within the virtual park, and σ represent the standard deviation of spatial distance for that virtual park. This embodiment employs an exponential decay function as the penalty mechanism, causing the penalty strength to increase non-linearly with the degree of spatial dispersion. The penalty is weak when the standard deviation of spatial distance is small, but rapidly decays once the standard deviation exceeds a reasonable range, thus enhancing the sensitivity of screening effective virtual parks.

[0052] As described above, by introducing the overall building confidence level and the number of associated enterprises as positive factors, the credibility evaluation of the virtual park not only depends on spatial location, but also fully considers the authority of the data source and the scale of enterprise agglomeration, avoiding the artificially high confidence level caused by relying solely on spatial distance. By introducing the standard deviation of spatial distance as a penalty term, loose and non-compact building clusters cannot obtain high confidence levels, effectively suppressing low-quality virtual parks caused by accidental noise or boundary anomalies. This results in the final generated virtual park possessing the characteristics of reliable data, enterprise agglomeration, and compact space, providing a highly credible target unit for subsequent enterprise attribution.

[0053] S6 assigns each unmatched enterprise in the enterprise set to the corresponding virtual park through the associated building, and obtains the association confidence between each unmatched enterprise and the associated virtual park based on the matching confidence and the overall park confidence.

[0054] In one specific embodiment, S6 includes the following steps: For any unmatched enterprise, the product of the matching confidence between the unmatched enterprise and the associated building and the overall confidence of the corresponding virtual park is used as the association confidence between the unmatched enterprise and the associated virtual park.

[0055] The process involves transferring unmatched companies to virtual parks through intermediary buildings, and then multiplying the matching confidence between the unmatched company and the associated building with the overall confidence of the virtual park to obtain the association confidence between the unmatched company and the associated virtual park. This association confidence is used to measure the overall reliability of the company's affiliation to the virtual park.

[0056] As described above, by assigning unmatched enterprises to corresponding virtual parks through their associated buildings, unmatched enterprises do not need to directly perform complex spatial calculations with virtual parks. Instead, they can achieve indirect assignment by leveraging existing unmatched enterprise-building associations and building-virtual park attribution relationships, simplifying the matching logic and improving computational efficiency. By using the product of matching confidence and park comprehensive confidence as the association confidence, the reliability of the association between unmatched enterprises and virtual parks is simultaneously constrained by both the confidence of the distance between the enterprise and the building and the quality of the virtual park itself. Only when both are high will the association confidence be high, avoiding misjudgments caused by an artificially high single dimension.

[0057] In one specific implementation, the method for constructing a virtual campus and matching enterprises further includes the following steps: S7. For each virtual park, determine the target generation strategy from the preset boundary generation strategy based on the number of buildings in the virtual park, and obtain the park boundary of the virtual park according to the target generation strategy.

[0058] S8 associates and stores each virtual park, the park boundary and overall confidence level of each virtual park, the building identifier set of all buildings constituting each virtual park, and the enterprise identifier set of all unmatched enterprises belonging to each virtual park, to generate a virtual park dataset.

[0059] In this process, a differentiated geometric construction method is used to generate the boundary of the virtual park based on the number of buildings in the virtual park. This ensures that the boundary reflects the actual distribution of the buildings and has regular geometric features, so as to guarantee that the boundary shape of virtual parks of different sizes is reasonable and standardized.

[0060] In one specific embodiment, S7 includes the following steps: If the number of buildings is greater than or equal to 3, the first preset generation strategy is determined as the target generation strategy. The first preset generation strategy is: generate an outer envelope polygon based on the coordinates of the center points of all buildings corresponding to the virtual park, and extend the outer envelope polygon outward by a first preset distance to obtain the park boundary of the virtual park.

[0061] If the number of buildings in the virtual park is equal to 2, then the second preset generation strategy is determined as the target generation strategy. The second preset generation strategy is: to generate an elliptical boundary with the line connecting the center points of the two buildings as the major axis and the second preset distance as the minor axis radius along the direction perpendicular to the line connecting the two buildings, so as to obtain the park boundary of the virtual park.

[0062] If the number of buildings is greater than or equal to 3, the cluster formed by multiple buildings usually has an irregular outer contour. In this embodiment, the coordinates of the center points of all buildings in the virtual park are extracted to form a point set. The convex hull polygon of the point set is calculated using convex hull algorithms such as Graham scan or Andrew algorithm to outline the outer contour of the building cluster and fit the distribution of the outermost buildings. Then, each edge of the convex hull is moved outward by a first preset distance to form a new boundary, which can smooth the sharp edges of the convex hull and cover the gaps between buildings, while providing a certain transition area so that the park boundary is more in line with the actual park area.

[0063] The specific value of the first preset distance can be set by the implementer according to the actual situation. For example, in this embodiment, the first preset distance is 20 meters.

[0064] Two buildings cannot be constructed as convex polygons, but ellipses can be used as boundaries. Correspondingly, the ellipse has the line connecting the two buildings as its major axis, ensuring that both buildings are located inside the ellipse; and an elliptical boundary is generated along the direction perpendicular to the line connecting the two buildings with a second preset distance as its minor axis radius, forming a reasonable area coverage.

[0065] The second preset distance determines the coverage width perpendicular to the line connecting the two buildings, and its specific value can be set as k times the distance between the two buildings. Where k≥0.5, in this embodiment, k=1.

[0066] The above-mentioned differentiated boundary generation strategy based on the number of buildings in the virtual park enables the boundary shape to adaptively match the spatial distribution characteristics of building clusters of different sizes, thereby improving the rationality of the park boundary.

[0067] This embodiment establishes the association between unmatched enterprises and buildings based on the spatial distance between the enterprise center point and the building center point, and calculates the matching confidence score, providing high-quality, weighted spatial units and reliable basic weights for the construction of virtual parks. Virtual parks are generated by spatially clustering candidate building nodes, and the overall park confidence score is calculated by integrating the building's overall confidence score, the number of enterprises associated with the building, and the standard deviation of the spatial distance between buildings. This ensures that the generation of virtual parks considers both the credibility of the data source and the positive contribution of the enterprise clustering scale, while also introducing a spatial compactness penalty mechanism. This achieves a multi-dimensional quantitative evaluation of the virtual park's quality, avoiding the problems of loose clustering and inflated confidence scores caused by relying solely on spatial distance. By combining the matching confidence score and the overall park confidence score to obtain the association confidence score between unmatched enterprises and virtual parks, the enterprise attribution results comprehensively reflect the credibility of the matching between unmatched enterprises and buildings, as well as the spatial clustering effectiveness of the virtual park itself, improving the accuracy and reliability of enterprise spatial attribution.

[0068] Example 2 Embodiment 2 of the present invention provides a non-transitory computer-readable storage medium, which can be disposed in an electronic device to store at least one instruction or at least one program related to implementing a method in the method embodiment. The at least one instruction or at least one program is loaded and executed by the processor to implement the virtual park construction and enterprise matching method provided in the above embodiment.

[0069] Example 3 Embodiment 3 of the present invention provides an electronic device, which includes a processor and the non-transitory computer-readable storage medium of Embodiment 2 of the present invention.

[0070] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for constructing a virtual industrial park and matching enterprises, characterized in that, The method includes the following steps: S1, obtain a set of enterprises that have not been matched with a physical park, a building dataset from several data sources, and physical park data. Each unmatched enterprise in the enterprise set corresponds to the coordinates of the enterprise's center point. The building dataset includes several buildings, each of which corresponds to the coordinates of the building's center point and the building's overall confidence level. The physical park data includes the building information covered by each physical park. S2, based on the spatial distance between the center point coordinates of each enterprise and the center point coordinates of each building, associate each unmatched enterprise with the building with the closest spatial distance, and calculate the matching confidence between the unmatched enterprise and the associated building based on the spatial distance; S3, Filter out at least one building in the building dataset that is associated with an unmatched enterprise and is not covered by a physical park, as candidate building nodes; S4. Cluster the candidate building nodes according to the coordinates of the building center point to generate at least one building cluster, and determine each building cluster as a virtual park; S5. For each virtual park, the overall confidence level of the virtual park is calculated based on the overall confidence level of each building in the virtual park, the number of unmatched enterprises associated with each building, and the standard deviation of the spatial distance between any two buildings in the virtual park. The standard deviation of the spatial distance is used to characterize the tightness of the spatial distribution between buildings. S6, assign each unmatched enterprise in the enterprise set to the corresponding virtual park through the associated building, and obtain the association confidence between each unmatched enterprise and the associated virtual park based on the matching confidence and the overall park confidence.

2. The virtual park construction and enterprise matching method according to claim 1, characterized in that, The data sources include official registration data, map service provider POI data, and OSM open-source building outline data. S1 includes the following steps: Obtain the initial confidence level for each data source; For any given building, the overall building confidence level is calculated using a probability complementary cumulative multiplication fusion formula based on several data sources corresponding to the building and the initial confidence level of each data source.

3. The virtual park construction and enterprise matching method according to claim 1, characterized in that, S2 includes the following steps: Based on the nearest spatial distance and the preset baseline attenuation distance, the matching confidence between the unmatched enterprise and the associated building is calculated using an exponential decay function.

4. The virtual park construction and enterprise matching method according to claim 1, characterized in that, S2 also includes the following steps: If the matching confidence level is less than a preset confidence threshold, the association between the unmatched enterprise and the associated building is determined to be invalid, and the unmatched enterprise is removed from the enterprise set.

5. The virtual park construction and enterprise matching method according to claim 1, characterized in that, The overall confidence level of the park is positively correlated with the overall confidence level of the buildings and the number of unmatched enterprises, and negatively correlated with the standard deviation of spatial distance.

6. The virtual park construction and enterprise matching method according to claim 1, characterized in that, S6 includes the following steps: For any unmatched enterprise, the product of the matching confidence between the unmatched enterprise and the associated building and the overall confidence of the corresponding virtual park is used as the association confidence between the unmatched enterprise and the associated virtual park.

7. The virtual park construction and enterprise matching method according to claim 1, characterized in that, The method for constructing virtual parks and matching enterprises also includes the following steps: S7. For each virtual park, based on the number of buildings in the virtual park, determine the target generation strategy from the preset boundary generation strategy, and obtain the park boundary of the virtual park according to the target generation strategy. S8 associates and stores each virtual park, the park boundary and overall confidence level of each virtual park, the building identifier set of all buildings constituting each virtual park, and the enterprise identifier set of all unmatched enterprises belonging to each virtual park, to generate a virtual park dataset.

8. The virtual park construction and enterprise matching method according to claim 7, characterized in that, S7 includes the following steps: If the number of buildings is greater than or equal to 3, the first preset generation strategy is determined as the target generation strategy. The first preset generation strategy is: to generate an outer envelope polygon based on the coordinates of the center points of all buildings corresponding to the virtual park, and to extend the outer envelope polygon outward by a first preset distance to obtain the park boundary of the virtual park. If the number of buildings in the virtual park is equal to 2, then the second preset generation strategy is determined as the target generation strategy. The second preset generation strategy is: to generate an elliptical boundary with the line connecting the center points of the two buildings as the major axis and the second preset distance as the minor axis radius along the direction perpendicular to the line connecting the two buildings, so as to obtain the park boundary of the virtual park.

9. A non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores at least one instruction or at least one program segment, characterized in that, The at least one instruction or the at least one program segment is loaded and executed by the processor to implement the virtual campus construction and enterprise matching method as described in any one of claims 1-8.

10. An electronic device, characterized in that, Includes a processor and the non-transitory computer-readable storage medium as described in claim 9.

Images