A multi-level POI semantic clustering method and system for large-screen maps

Abstract

This invention discloses a multi-level POI semantic clustering method and system for large-screen maps, belonging to the field of map data processing technology. The method includes the following steps: constructing a POI semantic tagging system, including POI category subdivisions and corresponding semantic tags, and establishing semantic relationships between different POI categories; dividing the map display levels into macro, meso, and micro levels according to the scaling ratio of the large-screen map, with each level corresponding to different clustering granularity requirements; performing POI semantic clustering based on the constructed semantic tagging system at different map display levels; and updating the clustering results when new POI data is added, existing POI data changes, or the map display level changes. This invention achieves accurate clustering, adapts to multi-level displays, ensures efficient real-time updates, and is suitable for various large-screen map application scenarios, improving the logic, readability, and user experience of map information display.

Description

Technical Field

[0001] This invention relates to the field of map data processing technology, specifically a multi-level POI semantic clustering method and system for large-screen maps. Background Technology

[0002] In large-screen map applications, Points of Interest (POI) data is abundant, diverse, and massive. Traditional POI clustering methods often focus on spatial aggregation, clustering based solely on spatial factors such as distance between POIs, neglecting the semantic information they contain. This results in clustering results lacking semantic logic. For example, grouping gas stations and hospitals, which are close to each other but have completely different functions, into one category fails to clearly convey information when displayed on the map, making it difficult for users to quickly find the information they need. Furthermore, existing clustering methods cannot flexibly adjust to the display level of the large-screen map (such as different zoom levels). During map zooming, the clustering results are either too coarse to display detailed information or too cluttered, with numerous dots, affecting the overall readability and display efficiency of the map, failing to meet users' demands for accurate and real-time information display on large-screen maps. Therefore, there is an urgent need for a method and device that can consider the semantic information of POIs and achieve multi-level clustering display. Summary of the Invention

[0003] The technical objective of this invention is to address the above-mentioned shortcomings by providing a multi-level POI semantic clustering method and system for large-screen maps. This method enables accurate clustering, adapts to multi-level displays, ensures efficient real-time updates, and is applicable to various large-screen map application scenarios, significantly improving the logic, readability, and user experience of map information display.

[0004] The technical solution adopted by this invention to solve its technical problem is:

[0005] A multi-level POI semantic clustering method for large-screen maps, the implementation of which includes the following steps:

[0006] Construct a multi-level, multi-dimensional POI semantic tagging system, which includes POI category subdivisions and corresponding semantic tags, and establishes semantic associations between different POI categories;

[0007] Based on the scaling ratio of the large-screen map, the map display levels are divided into macro level, meso level and micro level, each level corresponding to different clustering granularity requirements; the macro level focuses on the overall distribution of POIs in the region, with a coarser clustering granularity; the micro level needs to display detailed POI information, with a finer clustering granularity.

[0008] POI semantic clustering is performed based on the constructed semantic tagging system at different map display levels:

[0009] At the macro level, based on the macro semantic category and regional distribution characteristics of POIs, POIs with similar macro semantic categories and spatially adjacent are clustered into one category, and the central location of the region is used as the cluster point, highlighting semantic attributes in the cluster point label.

[0010] At the meso level, based on the macro clustering results, a second clustering is performed on POIs under the same macro category according to the meso semantic labels and precise spatial location relationships of POIs, making the clustering results more hierarchical.

[0011] At the micro level, based on the detailed semantic labels and precise spatial locations of POIs, POIs are clustered in detail, and POIs with close semantic associations and spatial proximity are grouped together to display their detailed semantic information and their association with surrounding related POIs.

[0012] A real-time data monitoring module is established. When new POI data is added, existing POI data changes, or the map display level changes, an incremental calculation method is used to recalculate and adjust the affected POIs and their clusters, and update the clustering results.

[0013] This method first constructs a multi-layered and multi-dimensional POI semantic tagging system, dividing the large-screen map display into macro, meso, and micro levels. At different levels, POI semantic clustering is performed based on the semantic tagging system. The macro level focuses on summarizing the regional POI distribution, the meso level refines macro clustering, and the micro level achieves detailed clustering. A dynamic update mechanism is also included, employing incremental computation to respond promptly to data changes and level switching, updating the clustering results accordingly. By introducing semantic information, accurate clustering is achieved, adapting to multi-level displays, ensuring efficient real-time updates, and making it suitable for various large-screen map application scenarios. This significantly improves the logicality, readability, and user experience of map information display.

[0014] Furthermore, the construction of the POI semantic tagging system,

[0015] Collect detailed information on various POIs, including name, function, target audience, and business hours. Based on natural language processing technology and domain knowledge, POI categories are subdivided and assigned corresponding semantic tags. Semantic relationships are established by analyzing business relationships and user behavior data between POIs.

[0016] Furthermore, the construction of the POI semantic tagging system is specifically implemented as follows:

[0017] Using web crawling technology, POI data is collected from data sources including various map service platforms and industry databases. After data cleaning and preprocessing, word vector models in natural language processing (such as Word2Vec) are used to perform semantic analysis on the text information of POIs, including names and descriptions, to extract keywords and semantic features.

[0018] Based on the pre-defined semantic label classification rules, POIs are classified into corresponding semantic categories, and each POI is assigned a specific semantic label.

[0019] At the same time, by analyzing the business relationships and user behavior data between POIs, semantic associations between POIs are established, ultimately forming a complete POI semantic tagging system.

[0020] Furthermore, in large-screen map applications, different zoom level thresholds can be set to divide the display hierarchy:

[0021] When the map zoom level is less than 1:100000, it is determined to be a macro level;

[0022] When the scaling ratio is between 1:100000 and 1:10000, it is considered a meso-level.

[0023] When the scaling ratio is greater than 1:10000, it is a micro level;

[0024] During map display, the current zoom level is obtained in real time, and the display level of the map is determined by comparing the zoom level with a preset threshold.

[0025] Furthermore, in the step of macroscopic hierarchical clustering, a density-based clustering algorithm is used, which combines the spatial location relationship of POIs to perform clustering operations.

[0026] At the macro level, the map is divided into several large regional units. For each regional unit, POIs are initially filtered based on their macro semantic category, extracting POIs of the same category (e.g., "residential area," "commercial area," "industrial area," etc.). Then, a density-based clustering algorithm (such as the DBSCAN algorithm) is used, combined with the spatial relationships of the POIs, to cluster spatially adjacent POIs belonging to the same macro semantic category into one cluster. The center of each cluster is calculated as the representative point of that cluster, and displayed on the map with corresponding icons and text labels, including the macro semantic category of the cluster and the number of POIs.

[0027] Furthermore, in the step of meso-level clustering, a hierarchical clustering algorithm is used to merge and cluster POIs according to semantic similarity and spatial distance.

[0028] When the meso-level clustering is triggered, the results of the macro-level clustering are further analyzed to understand the meso-level semantic labels and spatial relationships of POIs within each cluster. For example, for the "commercial area" cluster, POIs of the same meso-level semantic category are clustered again based on their meso-level semantic labels (such as "shopping mall," "supermarket," "restaurant," etc.) and more precise spatial distance. A hierarchical clustering algorithm is used to gradually merge POIs according to semantic similarity and spatial distance, forming the meso-level clustering results. The results are then updated and displayed on the map, distinguishing different meso-level clusters with different colors and icons, and displaying more detailed semantic information in the cluster point markers.

[0029] When the map is zoomed to a micro level, the most detailed clustering analysis is performed on the clustering results from the meso level or on individual POIs that are not clustered. Based on the detailed semantic labels and precise spatial locations of the POIs, POIs with close semantic associations and spatial proximity are grouped together. At this point, the clustering granularity can be refined to specific shops or facilities, and each cluster unit can be a single POI or a combination of a few closely related POIs. The clustering results are displayed on the map with high-resolution icons and detailed text descriptions. Users can click on cluster points to view detailed information about the POI, surrounding related POIs, and related semantic explanations.

[0030] Furthermore, the real-time data monitoring module is specifically implemented as follows:

[0031] A real-time data monitoring thread runs in the background to periodically check if there are any new data updates in the POI data source, and at the same time listens for changes in the map display hierarchy.

[0032] When new POI data is detected, the location and semantic information of the new POI are obtained to determine its map region and display level;

[0033] If the new POI is within the current display level, it will be included in the corresponding cluster unit, the center position and related attributes of the cluster will be recalculated, and the map display will be updated.

[0034] For changes to existing POI data (such as location movement or semantic attribute modification), first remove the POI from the original cluster, then reassign it to a suitable cluster based on the updated information, and adjust the affected clusters.

[0035] When the map display level changes, the clustering calculation for the current level is immediately stopped, and the semantic clustering algorithm for the corresponding level is re-executed according to the new level requirements to quickly generate new clustering results and update the map display.

[0036] This invention also claims a multi-level POI semantic clustering system for large-screen maps, comprising:

[0037] The semantic tagging system construction module is used to build a multi-level, multi-dimensional POI semantic tagging system. This system includes POI category subdivisions and corresponding semantic tags, and establishes semantic relationships between different POI categories.

[0038] The display hierarchy division module is used to divide the map display hierarchy into macro level, meso level and micro level according to the scaling ratio of the large screen map. Each level corresponds to different clustering granularity requirements.

[0039] The multi-level semantic clustering module is used to perform POI semantic clustering based on the constructed semantic tag system at different map display levels;

[0040] The dynamic update module is used to establish real-time data monitoring. When new POI data is added, existing POI data changes, or the map display level changes, an incremental calculation method is used to recalculate and adjust the affected POIs and their clusters, and update the clustering results.

[0041] The system achieves multi-level POI semantic clustering for large-screen maps using the methods described above.

[0042] The present invention also claims a multi-level POI semantic clustering implementation device for large-screen maps, comprising: at least one memory and at least one processor;

[0043] The at least one memory is used to store a machine-readable program;

[0044] The at least one processor is used to call the machine-readable program to implement the above method.

[0045] The present invention also claims a computer-readable medium storing computer instructions that, when executed by a processor, implement the above-described method.

[0046] Compared with existing technologies, the multi-level POI semantic clustering method and system for large-screen maps of the present invention have the following advantages:

[0047] 1. Precise semantic clustering: By constructing a multi-level semantic tagging system, the semantic information and relationships of POIs are fully considered, achieving more precise clustering that meets the user's cognitive and usage needs, making the display of POI information on the large-screen map more logical and readable.

[0048] 2. Adapt to multi-level display: Dynamic clustering is performed based on different display levels of the large-screen map. An overall overview is provided at the macro level, while detailed information is displayed at the micro level, which meets the user's needs for map information in different scenarios and improves the user experience.

[0049] 3. Efficient and real-time updates: Employing incremental calculation and dynamic update mechanisms, it can quickly respond to data changes and map level switching, ensuring that the information displayed on the map is always up-to-date and accurate, thus improving the practicality and reliability of map applications.

[0050] 4. Wide applicability: The method and system of this invention are applicable to various types of large-screen map application scenarios, such as urban planning display, tourist guide screens, traffic monitoring screens, etc., and have strong versatility and scalability. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating the multi-level POI semantic clustering method for large-screen maps provided in this embodiment of the invention. Detailed Implementation

[0052] This invention provides a multi-level POI semantic clustering method for large-screen maps. The method includes semantic tag system construction, display hierarchy division, multi-level semantic clustering process, and dynamic update mechanism. Its implementation includes the following steps:

[0053] Construct a multi-level, multi-dimensional POI semantic tagging system, which includes POI category subdivisions and corresponding semantic tags, and establishes semantic relationships between different POI categories.

[0054] Based on the scaling ratio of the large-screen map, the map display levels are divided into macro level, meso level and micro level, and each level corresponds to different clustering granularity requirements;

[0055] POI semantic clustering is performed based on the constructed semantic tagging system at different map display levels:

[0056] At the macro level, based on the macro semantic category and regional distribution characteristics of POIs, POIs with similar macro semantic categories and spatially adjacent are clustered into one category, and the central location of the region is used as the cluster point, highlighting semantic attributes in the cluster point label.

[0057] At the meso level, based on the macro clustering results, a second clustering is performed on POIs under the same macro category according to the meso semantic labels and precise spatial location relationships of POIs, making the clustering results more hierarchical.

[0058] At the micro level, based on the detailed semantic labels and precise spatial locations of POIs, POIs are clustered in detail, and POIs with close semantic associations and spatial proximity are grouped together to display their detailed semantic information and their association with surrounding related POIs.

[0059] A real-time data monitoring module is established. When new POI data is added, existing POI data changes, or the map display level changes, an incremental calculation method is used to recalculate and adjust the affected POIs and their clusters, and update the clustering results.

[0060] The semantic tagging system is constructed by collecting detailed information on various POIs, including name, function, target audience, and operating hours. Based on natural language processing technology and domain knowledge, a multi-level and multi-dimensional semantic tagging system for POIs is built. For example, the POI category of "hospital" is subdivided into subcategories such as "general hospital," "specialized hospital," and "community hospital," and each subcategory is assigned a corresponding semantic tag. For example, the semantic tags for "general hospital" may include "comprehensive medical services" and "complete departments." At the same time, semantic associations between different POI categories are established, such as the association between "pharmacies" and "hospitals" as "auxiliary medical care."

[0061] The display hierarchy is divided into multiple levels based on the zoom level of the large-screen map: macro level (low zoom level, displaying a large area), meso level (medium zoom level), and micro level (high zoom level, displaying a small, detailed area). Each level corresponds to a different clustering granularity requirement. The macro level focuses on summarizing the overall POI distribution of the region, with a coarser clustering granularity; the micro level needs to display detailed POI information, with a finer clustering granularity.

[0062] The multi-level semantic clustering process is as follows: At the macro level, priority is given to the macro semantic categories and regional distribution characteristics of POIs. POIs with similar macro semantic categories and spatial proximity are clustered together. For example, all commercial area-related POIs (shopping malls, shopping centers, etc.) in a city are aggregated together, with the center location of the area as the cluster point. The semantic attribute of "commercial area" is highlighted in the cluster point's label. As the map zooms to the meso level, the clustering is further refined based on the macro clustering. POIs within the same macro category are secondary-clustered according to their meso semantic labels and more precise spatial relationships. For example, "shopping malls" and "shopping centers" within a commercial area are clustered separately, making the clustering results more hierarchical. When the map is at the micro level, the most detailed clustering of POIs is performed based on their detailed semantic labels and precise spatial locations. Even a single POI can be treated as a cluster unit, simultaneously displaying its detailed semantic information and its association with surrounding related POIs.

[0063] The dynamic update mechanism automatically triggers cluster updates when new POI data is added, existing POI data changes (e.g., location shifts, semantic attribute updates), or the map display hierarchy changes. An incremental calculation method is used, recalculating and adjusting only the affected POIs and their respective clusters, avoiding a complete re-clustering of the entire map data. This ensures that the clustering results reflect the latest information promptly and accurately, improving the real-time performance of the map display.

[0064] The specific implementation of this method is as follows:

[0065] Step S1: Construct a semantic tag system.

[0066] POI data is collected from various map service platforms and industry databases using web crawling technology. After data cleaning and preprocessing, word vector models (such as Word2Vec) from natural language processing are used to perform semantic analysis on the textual information such as POI names and descriptions, extracting keywords and semantic features. Based on pre-defined semantic tagging rules, POIs are classified into corresponding semantic categories, and each POI is assigned a specific semantic tag. Simultaneously, by analyzing business relationships and user behavior data between POIs, semantic associations between POIs are established, ultimately forming a complete POI semantic tagging system.

[0067] Step S2: Divide the map display layers.

[0068] In large-screen map applications, different zoom level thresholds are set to classify display layers. For example, when the map zoom level is less than 1:100,000, it is considered a macro level; when the zoom level is between 1:100,000 and 1:10,000, it is considered a meso level; and when the zoom level is greater than 1:10,000, it is considered a micro level. During map display, the current zoom level is obtained in real time, and the display layer of the map is determined by comparing the zoom level with the preset threshold.

[0069] Step S3: Multi-level semantic clustering.

[0070] Macro-level clustering: At the macro level, the map is divided into several large regional units. For POIs within each regional unit, they are initially filtered based on their macro-semantic categories, extracting POIs of the same macro-semantic category (e.g., "residential area," "commercial area," "industrial area," etc.). Then, a density-based clustering algorithm (such as the DBSCAN algorithm) is used, combined with the spatial location relationships of POIs, to cluster spatially adjacent POIs belonging to the same macro-semantic category into one cluster. The center position of each cluster is calculated as the representative point of that cluster, and displayed on the map with corresponding icons and text labels, including the macro-semantic category of the cluster and the number of POIs.

[0071] Meso-level clustering: When triggered at the meso-level, the results of macro-level clustering are further analyzed to understand the meso-level semantic labels and spatial relationships of POIs within each cluster. For example, for the "commercial area" cluster, POIs of the same meso-level semantic category are clustered again based on their meso-level semantic labels (such as "shopping mall," "supermarket," "restaurant," etc.) and more precise spatial distance. A hierarchical clustering algorithm is used to gradually merge POIs according to semantic similarity and spatial distance, forming meso-level clustering results. These results are updated and displayed on the map, distinguishing different meso-level clusters with different colors and icons, and displaying more detailed semantic information in the cluster point markers.

[0072] Micro-level clustering: When the map is zoomed to the micro-level, the most detailed clustering analysis is performed on the clustering results of the meso-level or individual POIs that are not clustered. Based on the detailed semantic labels and precise spatial locations of the POIs, POIs with close semantic associations and spatial proximity are clustered together. At this point, the clustering granularity can be refined to specific shops or facilities, and each cluster unit can be a single POI or a combination of a few closely related POIs. The clustering results are displayed on the map with high-resolution icons and detailed text descriptions. Users can click on cluster points to view detailed information about the POI, surrounding related POIs, and related semantic explanations.

[0073] Step S4: Dynamic update.

[0074] A real-time data monitoring thread runs in the background, periodically checking for new data updates to the POI data source and listening for changes in map display hierarchy. When new POI data is detected, its location and semantic information are obtained to determine its map region and display hierarchy. If it falls within the current display hierarchy, the new POI is included in the corresponding cluster, its cluster center and related attributes are recalculated, and the map display is updated. For changes to existing POI data (such as location movement or semantic attribute modification), the POI is first removed from its original cluster, then reassigned to a suitable cluster based on the updated information, and the affected clusters are adjusted. When the map display hierarchy changes, cluster calculations for the current hierarchy are immediately stopped, and the corresponding semantic clustering algorithm is re-executed according to the new hierarchy requirements to quickly generate new clustering results and update the map display.

[0075] This method first constructs a multi-layered and multi-dimensional POI semantic tagging system, dividing the large-screen map display into macro, meso, and micro levels. At different levels, POI semantic clustering is performed based on the semantic tagging system. The macro level focuses on summarizing the regional POI distribution, the meso level refines macro clustering, and the micro level achieves detailed clustering. A dynamic update mechanism is also included, employing incremental computation to respond promptly to data changes and level switching, updating the clustering results accordingly. By introducing semantic information, accurate clustering is achieved, adapting to multi-level displays, ensuring efficient real-time updates, and making it suitable for various large-screen map application scenarios. This significantly improves the logicality, readability, and user experience of map information display.

[0076] This invention also provides a multi-level POI semantic clustering system for large-screen maps, including:

[0077] 1. Semantic Tag System Construction Module: This module is used to build a multi-level, multi-dimensional POI semantic tag system. This system includes POI category subdivisions and corresponding semantic tags, while also establishing semantic relationships between different POI categories. The implementation is as follows:

[0078] Detailed information on various Points of Interest (POIs) is collected, including name, function, target audience, and operating hours. Based on natural language processing technology and domain knowledge, a multi-level, multi-dimensional semantic tagging system for POIs is constructed. For example, the POI category of "hospital" is subdivided into subcategories such as "general hospital," "specialty hospital," and "community hospital," and each subcategory is assigned a corresponding semantic tag. For instance, the semantic tags for "general hospital" could include "comprehensive medical services" and "complete departments." Simultaneously, semantic relationships between different POI categories are established; for example, "pharmacies" and "hospitals" are associated with "auxiliary medical care."

[0079] 2. Display hierarchy segmentation module: This module divides the map display hierarchy into macro level (low zoom level, displaying a large area), meso level (medium zoom level), and micro level (high zoom level, displaying a small, detailed area) based on the zoom ratio of the large-screen map. Each level corresponds to different clustering granularity requirements. The macro level focuses on summarizing the overall POI distribution of the region, with a coarser clustering granularity; the micro level needs to display detailed POI information, with a finer clustering granularity.

[0080] 3. Multi-level semantic clustering module, used to perform POI semantic clustering based on the constructed semantic tag system at different map display levels.

[0081] At the macro level, priority is given to the macro semantic category and regional distribution characteristics of POIs. POIs with similar macro semantic categories and spatially adjacent are clustered into one category. For example, all commercial area-related POIs (shopping malls, shopping centers, etc.) in the city are aggregated together, and the central location of the area is used as the cluster point. At the same time, the semantic attribute of "commercial area" is highlighted in the label of the cluster point.

[0082] As the map zooms in to the meso-level, the clustering is further refined based on the macro-level clustering. At the meso-level, based on the macro-level clustering results, POIs under the same macro-level category are clustered again according to their meso-level semantic labels and precise spatial location relationships. For example, "shopping malls" and "shopping centers" in a commercial area are clustered separately, making the clustering results more hierarchical.

[0083] At the micro level, based on the detailed semantic labels and precise spatial locations of POIs, the most detailed clustering of POIs is performed, grouping POIs with close semantic associations and spatial proximity into one class. It is even possible to treat a single POI as a clustering unit, displaying its detailed semantic information and its association with surrounding related POIs.

[0084] 4. Dynamic Update Module: When new POI data is added, existing POI data changes (e.g., location shift, semantic attribute update), or the map display hierarchy changes, the system automatically triggers cluster updates. An incremental calculation method is used, recalculating and adjusting only the affected POIs and their respective clusters, avoiding a complete re-clustering of the entire map data. This ensures that the clustering results reflect the latest information promptly and accurately, improving the real-time performance of the map display.

[0085] This system implements multi-level POI semantic clustering for large-screen maps using the multi-level POI semantic clustering method described in the above embodiments. The specific implementation is as follows:

[0086] S1, the semantic tag system construction module, implements the construction of the semantic tag system.

[0087] POI data is collected from various map service platforms and industry databases using web crawling technology. After data cleaning and preprocessing, word vector models (such as Word2Vec) from natural language processing are used to perform semantic analysis on the textual information such as POI names and descriptions, extracting keywords and semantic features. Based on pre-defined semantic tagging rules, POIs are classified into corresponding semantic categories, and each POI is assigned a specific semantic tag. Simultaneously, by analyzing business relationships and user behavior data between POIs, semantic associations between POIs are established, ultimately forming a complete POI semantic tagging system.

[0088] S2, The display hierarchy module divides the map display hierarchy.

[0089] In large-screen map applications, different zoom level thresholds are set to classify display layers. For example, when the map zoom level is less than 1:100,000, it is considered a macro level; when the zoom level is between 1:100,000 and 1:10,000, it is considered a meso level; and when the zoom level is greater than 1:10,000, it is considered a micro level. During map display, the current zoom level is obtained in real time, and the display layer of the map is determined by comparing the zoom level with the preset threshold.

[0090] S3, the multi-level semantic clustering module implements multi-level semantic clustering.

[0091] Macro-level clustering: At the macro level, the map is divided into several large regional units. For POIs within each regional unit, they are initially filtered based on their macro-semantic categories, extracting POIs of the same macro-semantic category (e.g., "residential area," "commercial area," "industrial area," etc.). Then, a density-based clustering algorithm (such as the DBSCAN algorithm) is used, combined with the spatial location relationships of POIs, to cluster spatially adjacent POIs belonging to the same macro-semantic category into one cluster. The center position of each cluster is calculated as the representative point of that cluster, and displayed on the map with corresponding icons and text labels, including the macro-semantic category of the cluster and the number of POIs.

[0092] Meso-level clustering: When triggered at the meso-level, the results of macro-level clustering are further analyzed to understand the meso-level semantic labels and spatial relationships of POIs within each cluster. For example, for the "commercial area" cluster, POIs of the same meso-level semantic category are clustered again based on their meso-level semantic labels (such as "shopping mall," "supermarket," "restaurant," etc.) and more precise spatial distance. A hierarchical clustering algorithm is used to gradually merge POIs according to semantic similarity and spatial distance, forming meso-level clustering results. These results are updated and displayed on the map, distinguishing different meso-level clusters with different colors and icons, and displaying more detailed semantic information in the cluster point markers.

[0093] Micro-level clustering: When the map is zoomed to the micro-level, the most detailed clustering analysis is performed on the clustering results of the meso-level or individual POIs that are not clustered. Based on the detailed semantic labels and precise spatial locations of the POIs, POIs with close semantic associations and spatial proximity are clustered together. At this point, the clustering granularity can be refined to specific shops or facilities, and each cluster unit can be a single POI or a combination of a few closely related POIs. The clustering results are displayed on the map with high-resolution icons and detailed text descriptions. Users can click on cluster points to view detailed information about the POI, surrounding related POIs, and related semantic explanations.

[0094] S4. The dynamic update module establishes real-time data monitoring and enables dynamic updates.

[0095] A real-time data monitoring thread runs in the background, periodically checking for new data updates to the POI data source and listening for changes in map display hierarchy. When new POI data is detected, its location and semantic information are obtained to determine its map region and display hierarchy. If it falls within the current display hierarchy, the new POI is included in the corresponding cluster, its cluster center and related attributes are recalculated, and the map display is updated. For changes to existing POI data (such as location movement or semantic attribute modification), the POI is first removed from its original cluster, then reassigned to a suitable cluster based on the updated information, and the affected clusters are adjusted. When the map display hierarchy changes, cluster calculations for the current hierarchy are immediately stopped, and the corresponding semantic clustering algorithm is re-executed according to the new hierarchy requirements to quickly generate new clustering results and update the map display.

[0096] This invention also provides a multi-level POI semantic clustering implementation device for large-screen maps, comprising: at least one memory and at least one processor;

[0097] The at least one memory is used to store a machine-readable program;

[0098] The at least one processor is used to call the machine-readable program to implement the multi-level POI semantic clustering method for large-screen maps described in the above embodiments.

[0099] This invention also provides a computer-readable medium storing computer instructions. When executed by a processor, the computer instructions cause the processor to perform the multi-level POI semantic clustering method for large-screen maps described in the above embodiments. Specifically, a system or device equipped with a storage medium storing software program code that implements the functions of any of the embodiments described above can be provided, and the computer (or CPU or MPU) of the system or device can read and execute the program code stored in the storage medium.

[0100] In this case, the program code read from the storage medium can itself implement the function of any of the above embodiments, and therefore the program code and the storage medium storing the program code constitute part of the present invention.

[0101] Examples of storage media used to provide program code include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, program code can be downloaded from a server computer via a communication network.

[0102] Furthermore, it should be clear that not only can the program code read by the computer be executed, but also the operating system or other components operating on the computer can be instructed based on the program code to perform some or all of the actual operations, thereby realizing the function of any of the embodiments described above.

[0103] Furthermore, it is understood that the program code read from the storage medium is written to the memory set in the expansion board inserted into the computer or to the memory set in the expansion unit connected to the computer. Then, based on the instructions of the program code, the CPU or other components installed on the expansion board or expansion unit execute some and all of the actual operations, thereby realizing the function of any of the embodiments described above.

[0104] The present invention has been shown and described in detail above with reference to the accompanying drawings and preferred embodiments. However, the present invention is not limited to these disclosed embodiments. Based on the above embodiments, those skilled in the art will know that more embodiments of the present invention can be obtained by combining the code review methods in the different embodiments. These embodiments are also within the protection scope of the present invention.

Claims

1. A multi-level POI semantic clustering method for large-screen maps, characterized in that, The implementation of this method includes the following steps: Construct a multi-level, multi-dimensional POI semantic tagging system, which includes POI category subdivisions and corresponding semantic tags, and establishes semantic associations between different POI categories; Based on the scaling ratio of the large-screen map, the map display levels are divided into macro level, meso level and micro level, and each level corresponds to different clustering granularity requirements; POI semantic clustering is performed based on the constructed semantic tagging system at different map display levels: At the macro level, based on the macro semantic category and regional distribution characteristics of POIs, POIs with similar macro semantic categories and spatially adjacent are clustered into one category, and the central location of the region is used as the cluster point, highlighting semantic attributes in the cluster point label. At the meso level, based on the macro clustering results, a second clustering is performed on POIs under the same macro category according to the meso semantic labels and precise spatial location relationships of POIs, making the clustering results more hierarchical. At the micro level, based on the detailed semantic labels and precise spatial locations of POIs, POIs are clustered in detail, and POIs with close semantic associations and spatial proximity are grouped together to display their detailed semantic information and their association with surrounding related POIs. A real-time data monitoring module is established. When new POI data is added, existing POI data changes, or the map display level changes, an incremental calculation method is used to recalculate and adjust the affected POIs and their clusters, and update the clustering results.

2. The multi-level POI semantic clustering method for large-screen maps according to claim 1, characterized in that, The construction of the POI semantic tagging system. Collect detailed information on various POIs, including name, function, target audience, and business hours. Based on natural language processing technology, POI categories are subdivided and assigned corresponding semantic tags. Semantic relationships are established by analyzing business relationships and user behavior data between POIs.

3. A multi-level POI semantic clustering method for large-screen maps according to claim 1 or 2, characterized in that, The specific implementation of the POI semantic tagging system is as follows: We use web crawling technology to collect POI data from various map service platforms and industry databases. After data cleaning and preprocessing, we use word vector models in natural language processing to perform semantic analysis on the text information of POIs, including names and descriptions, and extract keywords and semantic features. Based on pre-defined semantic label classification rules, POIs are classified into corresponding semantic categories, and each POI is assigned a specific semantic label. At the same time, by analyzing the business relationships and user behavior data between POIs, semantic associations between POIs are established, ultimately forming a complete POI semantic tagging system.

4. The multi-level POI semantic clustering method for large-screen maps according to claim 1, characterized in that, In large-screen map applications, different zoom level thresholds can be set to divide the display hierarchy: When the map zoom level is less than 1:100000, it is determined to be a macro level; When the scaling ratio is between 1:100000 and 1:10000, it is considered a meso-level. When the scaling ratio is greater than 1:10000, it is a micro level; During map display, the current zoom level is obtained in real time, and the display level of the map is determined by comparing the zoom level with a preset threshold.

5. A multi-level POI semantic clustering method for large-screen maps according to claim 1, characterized in that, In the macroscopic hierarchical clustering step, a density-based clustering algorithm is used, which combines the spatial location relationship of POIs to perform clustering operations.

6. The multi-level POI semantic clustering method for large-screen maps according to claim 1, characterized in that, In the step of clustering at the meso-level, a hierarchical clustering algorithm is used to merge and cluster POIs according to semantic similarity and spatial distance.

7. A multi-level POI semantic clustering method for large-screen maps according to claim 1, characterized in that, The real-time data monitoring module is implemented as follows: A real-time data monitoring thread runs in the background to periodically check if there are any new data updates in the POI data source, and at the same time listen for changes in the map display hierarchy. When new POI data is detected, the location and semantic information of the new POI are obtained to determine its map region and display level; If the new POI is within the current display level, it will be included in the corresponding cluster unit, the center position and related attributes of the cluster will be recalculated, and the map display will be updated. For changes to existing POI data, first remove the POI from the original cluster, then reassign it to a suitable cluster based on the updated information, and adjust the affected clusters. When the map display level changes, the clustering calculation for the current level is immediately stopped, and the semantic clustering algorithm for the corresponding level is re-executed according to the new level requirements to quickly generate new clustering results and update the map display.

8. A multi-level POI semantic clustering system for large-screen maps, characterized in that, include: The semantic tagging system construction module is used to build a multi-level, multi-dimensional POI semantic tagging system. This system includes POI category subdivisions and corresponding semantic tags, and establishes semantic relationships between different POI categories. The display hierarchy division module is used to divide the map display hierarchy into macro level, meso level and micro level according to the scaling ratio of the large screen map. Each level corresponds to different clustering granularity requirements. The multi-level semantic clustering module is used to perform POI semantic clustering based on the constructed semantic tag system at different map display levels; The dynamic update module is used to establish real-time data monitoring. When new POI data is added, existing POI data changes, or the map display level changes, an incremental calculation method is used to recalculate and adjust the affected POIs and their clusters, and update the clustering results. The system achieves multi-level POI semantic clustering for large-screen maps through the method described in any one of claims 1 to 7.

9. A multi-level POI semantic clustering implementation device for large-screen maps, characterized in that, include: At least one memory and at least one processor; The at least one memory is used to store a machine-readable program; The at least one processor is configured to invoke the machine-readable program to implement the method according to any one of claims 1 to 7.

10. A computer-readable medium, characterized in that, The computer-readable medium stores computer instructions that, when executed by a processor, implement the method described in any one of claims 1 to 7.

Images