Point of interest data processing method and apparatus, electronic device, and readable storage medium

Abstract

The application provides a point of interest data processing method and device, electronic equipment and readable storage medium, relates to the technical field of electronic map data processing, and the method comprises the following steps: supplementing a set of points of interest to be processed by using scene surface data, and performing deduplication processing to obtain a first set of points of interest; according to the spatial relationship of the points of interest, the spatial type of the points of interest in the first set of points of interest is divided to obtain at least one second set of points of interest; based on a pre-constructed map point of interest hierarchical relationship extraction model, address semantic units corresponding to address information of the point of interest and classification attributes corresponding to name information of the point of interest, the hierarchical coefficient of the point of interest in the second set of points of interest is determined. The scheme improves the data utilization rate, makes the display of the electronic map more reasonable, and improves the utilization rate of the electronic map.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of electronic map data processing, and particularly relates to a point of interest data processing method and device, electronic equipment and a readable storage medium. BACKGROUND

[0002] With the development of geospatial information technology, electronic maps, as a map form combining maps with computer technology, are more intuitive and diverse in displaying map information. The richness of map data and the distinctness of hierarchical structure are key factors affecting map display and user experience.

[0003] In the process of making electronic maps, the richness and reliability of POI (Point of Interest) data are key to measuring the efficiency of electronic map making, and the classification principle directly affects the universality of data. The more detailed the modification work is, the better the effect of the electronic map is. Currently, POI data is independent of other map data and does not combine with other data correlation information, so the data usage rate is not optimal. Moreover, the existing data de-duplication processing adopts manual interpretation for modification, so the data accuracy is low and the production cycle is long. In addition, there is no clear hierarchical relationship of POI data in the existing map data, so the sub-level POI points may be displayed before the parent-level POI points in the electronic map display, resulting in display errors and affecting the overall display effect of the electronic map and reducing the usage rate of the electronic map. SUMMARY

[0004] The purpose of the embodiments of the present application is to provide a point of interest data processing method and device, electronic equipment and a readable storage medium, so as to solve the problem that the data usage rate is not optimal and the hierarchical relationship of POI data is not clear in the prior art, affecting the display effect of the electronic map.

[0005] To achieve the above purpose, the embodiments of the present application provide a point of interest data processing method, comprising:

[0006] The scene surface data is used to supplement the point of interest set to be processed, and de-duplication processing is performed to obtain a first point of interest set;

[0007] According to the spatial relationship of the point of interest, the spatial type of the point of interest in the first point of interest set is divided to obtain at least one second point of interest set;

[0008] Based on the pre-constructed map point of interest hierarchical relationship extraction model, the address semantic unit corresponding to the address information of the point of interest and the classification attribute corresponding to the name information of the point of interest, the hierarchical coefficient of the point of interest in the second point of interest set is determined.

[0009] Optionally, the set of interest points to be processed is supplemented using scene surface data, and deduplication is performed to obtain a first set of interest points, including:

[0010] Spatial information, category information, and attribute information are extracted from the scene surface data, and the set of interest points corresponding to the scene surface data is obtained through dimensionality reduction processing.

[0011] Based on the set of interest points corresponding to the scene surface data, the set of interest points to be processed is supplemented to obtain a third set of interest points;

[0012] Extract the interest points of entities with the same name from the third set of interest points to obtain the fourth set of interest points;

[0013] According to preset rules, the fourth set of interest points is deduplicated to obtain the fifth set of interest points to be deleted.

[0014] The interest points corresponding to the fifth interest point set in the third interest point set are deleted to obtain the first interest point set.

[0015] Optionally, interest points of entities with the same name are extracted from the third set of interest points to obtain a fourth set of interest points, including:

[0016] Using the spatial, category, and attribute information of the interest point set corresponding to the scene surface data, the third interest point set is searched for category attribute similarity and spatial similarity, and interest points of entities with the same name in the third interest point set are extracted to obtain the fourth interest point set.

[0017] Among them, at least two points of interest whose name and address information meet at least one of the following criteria are considered to be points of interest of the same entity:

[0018] The name information is the same and the address information is the same;

[0019] The name and address information meet the first similarity condition;

[0020] The name and address information meet the second similarity condition;

[0021] The name and address information meet the second similarity condition.

[0022] Optionally, the preset rules include at least one of the following:

[0023] If at least two points of interest in the fourth set of points of interest have the same name information and the same address information, or if the name information and address information meet the first similarity condition, one point of interest is retained.

[0024] When the name and address information of at least two interest points in the fourth interest point set meet the second similarity condition, the interest point with the correct positional relationship with the scene plane is retained.

[0025] When the name and address information of at least two points of interest in the fourth set of points of interest meet the third similarity condition, the point of interest with the higher accuracy quality score is retained.

[0026] Optionally, it also includes:

[0027] Get the ratio of the number of semantic units in the name of a point of interest to the distance from that point of interest to the corresponding scene surface;

[0028] The accuracy quality score is obtained based on the ratio.

[0029] Optionally, the at least one set of second points of interest includes at least one of the following:

[0030] The point of interest is the first set of sub-points of interest located within the building;

[0031] The set of second sub-interest points located within the scene plane but not within buildings;

[0032] The set of third sub-interest points located within the scene plane and within buildings;

[0033] The fourth set of sub-interest points that are neither located within the scene plane nor within a building.

[0034] Optionally, it also includes:

[0035] Based on the building entity model in the pre-constructed geographic entity full-space model, points of interest are mapped to entity nodes to construct the map point of interest hierarchy relationship extraction model; wherein, the map point of interest hierarchy relationship extraction model includes a hierarchy coefficient table of points of interest.

[0036] Optionally, based on a pre-built map point of interest hierarchy extraction model, address semantic units corresponding to the address information of the points of interest, and classification attributes corresponding to the name information of the points of interest, the hierarchy coefficients of the points of interest in the second set of points of interest are determined, including:

[0037] The points of interest are mapped to the map point of interest hierarchy extraction model, and the hierarchy coefficient of the points of interest in the second set of points of interest is determined as a first preset value;

[0038] Based on the address semantic units corresponding to the address information of the points of interest, a structured analysis is performed to obtain the ending element, and based on the ending element, the hierarchical coefficient of the points of interest in the second set of points of interest is determined as the second preset value.

[0039] When the ending element includes a road address or directional word, the name information of the point of interest is segmented using a word segmentation engine to determine the semantic unit and semantic unit combination method corresponding to the name information. Based on the classification attribute of the semantic unit and the semantic unit combination method, the hierarchical coefficient of the point of interest in the second set of points of interest is determined to be a third preset value.

[0040] This invention also provides a point-of-interest (POI) data processing apparatus, comprising:

[0041] The processing module is used to supplement the set of interest points to be processed using scene surface data and perform deduplication to obtain the first set of interest points;

[0042] The partitioning module is used to partition the spatial types of the points of interest in the first set of points of interest according to the spatial relationships of the points of interest, so as to obtain at least one second set of points of interest.

[0043] The determination module is used to determine the hierarchy coefficients of the points of interest in the second set of points of interest based on a pre-built map point of interest hierarchy relationship extraction model, the address semantic units corresponding to the address information of the points of interest, and the classification attributes corresponding to the name information of the points of interest.

[0044] Optionally, the processing module includes:

[0045] The first processing unit is used to extract spatial information, category information and attribute information from the scene surface data, and obtain the set of interest points corresponding to the scene surface data through dimensionality reduction processing.

[0046] The supplementary unit is used to supplement the set of interest points to be processed based on the set of interest points corresponding to the scene surface data, so as to obtain a third set of interest points;

[0047] An extraction unit is used to extract the interest points of entities with the same name in the third set of interest points to obtain a fourth set of interest points.

[0048] The second processing unit is used to perform deduplication processing on the fourth set of interest points according to preset rules to obtain a fifth set of interest points to be deleted.

[0049] The deletion unit is used to delete the points of interest in the third set of points of interest that correspond to the fifth set of points of interest, thereby obtaining the first set of points of interest.

[0050] Optionally, the extraction unit is specifically used for:

[0051] Using the spatial, category, and attribute information of the interest point set corresponding to the scene surface data, the third interest point set is searched for category attribute similarity and spatial similarity, and interest points of entities with the same name in the third interest point set are extracted to obtain the fourth interest point set.

[0052] Among them, at least two points of interest whose name and address information meet at least one of the following criteria are considered to be points of interest of the same entity:

[0053] The name information is the same and the address information is the same;

[0054] The name and address information meet the first similarity condition;

[0055] The name and address information meet the second similarity condition;

[0056] The name and address information meet the third similarity condition.

[0057] Optionally, the preset rules include at least one of the following:

[0058] If at least two points of interest in the fourth set of points of interest have the same name information and the same address information, or if the name information and address information meet the first similarity condition, one point of interest is retained.

[0059] When the name and address information of at least two interest points in the fourth interest point set meet the second similarity condition, the interest point with the correct positional relationship with the scene plane is retained.

[0060] When the name and address information of at least two points of interest in the fourth set of points of interest meet the third similarity condition, the point of interest with the higher accuracy quality score is retained.

[0061] Optionally, the device further includes:

[0062] The acquisition module is used to obtain the ratio of the number of semantic units in the name of a point of interest to the distance from that point of interest to the corresponding scene surface;

[0063] The module is used to obtain the accuracy quality score based on the ratio.

[0064] Optionally, the at least one set of second points of interest includes at least one of the following:

[0065] The point of interest is the first set of sub-points of interest located within the building;

[0066] The set of second sub-interest points located within the scene plane but not within buildings;

[0067] The set of third sub-interest points located within the scene plane and within buildings;

[0068] The fourth set of sub-interest points that are neither located within the scene plane nor within a building.

[0069] Optionally, the device further includes:

[0070] The construction module is used to map points of interest to entity nodes based on the building entity models in the pre-built geographic entity full-space model, and to construct the map point of interest hierarchy relationship extraction model; wherein, the map point of interest hierarchy relationship extraction model includes a hierarchy coefficient table of points of interest.

[0071] Optionally, the determining module is specifically used for:

[0072] The points of interest are mapped to the map point of interest hierarchy extraction model, and the hierarchy coefficient of the points of interest in the second set of points of interest is determined as a first preset value;

[0073] Based on the address semantic units corresponding to the address information of the points of interest, a structured analysis is performed to obtain the ending element, and based on the ending element, the hierarchical coefficient of the points of interest in the second set of points of interest is determined as the second preset value.

[0074] When the ending element includes a road address or directional word, the name information of the point of interest is segmented using a word segmentation engine to determine the semantic unit and semantic unit combination method corresponding to the name information. Based on the classification attribute of the semantic unit and the semantic unit combination method, the hierarchical coefficient of the point of interest in the second set of points of interest is determined to be a third preset value.

[0075] This invention also provides an electronic device, including: a processor, a memory, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, it implements the steps of the point of interest data processing method as described in any of the preceding embodiments.

[0076] This invention also provides a readable storage medium storing a program that, when executed by a processor, implements the steps of the point of interest data processing method as described in any of the preceding embodiments.

[0077] The above-described technical solution of the present invention has at least the following beneficial effects:

[0078] In the above scheme, the set of points of interest to be processed is supplemented by scene surface data and deduplicated to obtain a first set of points of interest. Based on the spatial relationship of the points of interest, the spatial types of the points of interest in the first set of points of interest are divided to obtain at least one second set of points of interest. Based on the pre-constructed map point of interest hierarchy extraction model, the address semantic units corresponding to the address information of the points of interest, and the classification attributes corresponding to the name information of the points of interest, the hierarchy coefficients of the points of interest in the second set of points of interest are determined. This can make full use of the spatial relationship between map data, improve data utilization, and clarify the hierarchy coefficients of the points of interest, which facilitates the production of electronic maps, thereby improving the display effect of electronic maps and increasing the utilization rate of electronic maps. Attached Figure Description

[0079] Figure 1 This is one of the flowcharts illustrating the point-of-interest data processing method according to an embodiment of the present invention;

[0080] Figure 2 This is a second schematic flowchart of the point of interest data processing method according to an embodiment of the present invention;

[0081] Figure 3 This is one of the schematic diagrams illustrating the construction of a geographic entity full-space model for the point of interest data processing method according to an embodiment of the present invention;

[0082] Figure 4 This is the second schematic diagram illustrating the construction of a geographic entity full-space model for the point of interest data processing method according to an embodiment of the present invention.

[0083] Figure 5 This is a schematic diagram illustrating the construction of a map interest point hierarchy relationship extraction model for the interest point data processing method according to an embodiment of the present invention;

[0084] Figure 6 This is a third flowchart illustrating the point-of-interest data processing method according to an embodiment of the present invention;

[0085] Figure 7 This is one of the schematic diagrams of the point of interest data processing device according to an embodiment of the present invention;

[0086] Figure 8 This is a second schematic diagram of the point of interest data processing device according to an embodiment of the present invention;

[0087] Figure 9 This is a schematic diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0088] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0089] This invention addresses the problems in the prior art where data utilization is not optimal and the hierarchical relationship of POI data is unclear, affecting the display effect of electronic maps. It provides a method, apparatus, electronic device, and readable storage medium for processing point of interest data.

[0090] like Figure 1 As shown, an embodiment of the present invention provides a method for processing point-of-interest (POI) data, including:

[0091] Step 101: Use scene surface data to supplement the set of interest points to be processed and perform deduplication to obtain the first set of interest points;

[0092] It should be noted that the point of interest data processing method described in this embodiment of the invention merges POI data and scene surface data attribute information, enriches the point of interest set, and improves data utilization.

[0093] Step 102: Based on the spatial relationships of the points of interest, divide the spatial types of the points of interest in the first set of points of interest to obtain at least one second set of points of interest;

[0094] It should be noted that spatial relationships are used to divide the points of interest, resulting in at least one set of points of interest. Here, spatial relationships include: the spatial relationship between the points of interest and the scene plane, and the spatial relationship between the points of interest and buildings.

[0095] Step 103: Based on the pre-constructed map point of interest hierarchy relationship extraction model, the address semantic units corresponding to the address information of the points of interest, and the classification attributes corresponding to the name information of the points of interest, determine the hierarchy coefficients of the points of interest in the second set of points of interest.

[0096] It should be noted that the point of interest data processing method described in this embodiment of the invention clarifies the hierarchical relationship of points of interest in the map point of interest hierarchy extraction model, enriches the information in the electronic map database, makes the subsequent display effect of the electronic map more reasonable, and improves the utilization rate of the electronic map.

[0097] In this embodiment of the invention, a first set of points of interest (POIs) is obtained by supplementing the set of POIs to be processed with scene surface data and performing deduplication. Based on the spatial relationships of the POIs, the spatial types of the POIs in the first set of POIs are divided to obtain at least one second set of POIs. Based on a pre-constructed map POI hierarchy relationship extraction model, the address semantic units corresponding to the address information of the POIs, and the classification attributes corresponding to the name information of the POIs, the hierarchy coefficients of the POIs in the second set of POIs are determined. This fully utilizes the spatial relationships between map data, improves data utilization, and clarifies the hierarchy coefficients of the POIs, facilitating electronic map production, thereby improving the display effect of electronic maps and increasing their utilization rate.

[0098] Optionally, such as Figure 2 As shown, step 101: Use scene surface data to supplement the set of interest points to be processed and perform deduplication to obtain the first set of interest points, including:

[0099] Step 201: Extract spatial information, category information and attribute information from the scene surface data, and obtain the set of interest points N0 corresponding to the scene surface data through dimensionality reduction processing;

[0100] Step 202: Based on the set of interest points N0 corresponding to the scene surface data, supplement the set of interest points N to be processed to obtain a third set of interest points N1, where N1 = N∪N0;

[0101] Step 203: Extract the interest points of entities with the same name from the third interest point set N1 to obtain the fourth interest point set Q0, where Q0∈N1.

[0102] Step 204: According to the preset rules, the fourth set of interest points Q0 is deduplicated to obtain the fifth set of interest points Q1, where Q1∈Q0, where interest points need to be deleted.

[0103] Step 205: Delete the interest points in the third interest point set N1 that correspond to the fifth interest point set Q1 to obtain the first interest point set N2, where N2 = N1 - Q1.

[0104] It should be noted that the scene area data mainly represents functional areas such as residential areas, schools, hospitals, and scenic spots, possessing distinct spatial attributes, especially area map data. This embodiment of the invention utilizes the spatial and attribute correlation information between area map data and point-of-interest (POI) data to enrich the types of POIs and more comprehensively meet the application needs of the data.

[0105] Optionally, step 203: Extract interest points of entities with the same name from the third set of interest points to obtain a fourth set of interest points, including:

[0106] Using the spatial, category, and attribute information of the interest point set corresponding to the scene surface data, the third interest point set is searched for category attribute similarity and spatial similarity, and interest points of entities with the same name in the third interest point set are extracted to obtain the fourth interest point set.

[0107] Among them, at least two points of interest whose name and address information meet at least one of the following criteria are considered to be points of interest of the same entity:

[0108] The name information is the same and the address information is the same;

[0109] The name and address information meet the first similarity condition;

[0110] The name and address information meet the second similarity condition;

[0111] The name and address information meet the third similarity condition.

[0112] Here, the first similarity condition is that the name information is similar and the address information is the same; the second similarity condition is that the name information is the same and the address information is similar; the third similarity condition is that the name information is similar and the address information is different; specifically, the similarity judgment steps are as follows:

[0113] For example, given a specific scene area, a 1km buffer zone is established, and a set of interest points (POPs) with latitude and longitude coordinates located within this scene area and within a 1km radius is determined. The PPOs in this set are clustered, and points with the same category and attribute information as the scene area data are extracted. PPO matching is then performed based on category attribute similarity, i.e., non-spatial attribute similarity. Specifically, semantic similarity and character similarity methods are used to compare PPOs.

[0114] First, the name information field is divided into various semantic units based on semantic similarity. Then, character similarity matching is performed within each unit. Semantic similarity refers to matching similarity at the semantic level. For example, "What is your home address?" and "What is your home address?" have a high degree of semantic similarity.

[0115] SIM(A,B)=sim1(A(x),B(x))·sim2(A(x),B(x));

[0116] {A∈Q0, B∈Q0|SIM(A,B)>δ};

[0117] Where sim1(A(x), B(x)) is the semantic similarity of a semantic unit in two interest points A and B; sim2(A(x), B(x)) is the character similarity of a semantic unit in two interest points A and B; when SIM(A, B) is greater than the preset threshold δ, the two interest points A and B are determined to be interest points of the same entity.

[0118] Optionally, the preset rules include at least one of the following:

[0119] When at least two points of interest in the fourth set of points of interest have the same name information and the same address information, or when the name information and address information meet the first similarity condition, one point of interest is randomly retained.

[0120] When the name and address information of at least two interest points in the fourth interest point set meet the second similarity condition, the interest point with the correct positional relationship with the scene plane is retained.

[0121] When the name and address information of at least two points of interest in the fourth set of points of interest meet the third similarity condition, the point of interest with the higher accuracy quality score is retained.

[0122] It should be noted that, according to the preset rules described above, the interest points in the fourth interest point set Q0 of the same entity obtained in step 203 are deduplicated, including interest points with the same or similar name and address information, as well as interest points with the same name but different address information.

[0123] Optionally, it also includes:

[0124] Get the ratio of the number of semantic units in the name of a point of interest to the distance from that point of interest to the corresponding scene surface;

[0125] The accuracy quality score is obtained based on the ratio.

[0126] It should be noted that for points of interest with the same name but different address information, the following formula should be used for accuracy and quality assessment:

[0127]

[0128] Here, distance(interest point, scene plane) refers to the distance from the interest point to its corresponding scene plane. If both the interest point and the interest point being compared are located within the scene plane, the distance is 1. The score is determined by the number of semantic units in the name information field of the interest point, i.e. the information content of the interest point. Interest points with higher accuracy quality scores are retained to form a set.

[0129] Optionally, the at least one set of second points of interest includes at least one of the following:

[0130] The point of interest is located in the first sub-set of points of interest, P1, within the building.

[0131] The second sub-set of interest points P2, which are located within the scene plane but not within buildings;

[0132] The set of third sub-interest points P3, which are located within the scene plane and within buildings;

[0133] The fourth sub-set of interest points, P4, is located neither within the scene plane nor within a building.

[0134] Optionally, it also includes:

[0135] Based on the building entity model in the pre-constructed geographic entity full-space model, points of interest are mapped to entity nodes to construct the map point of interest hierarchy relationship extraction model; wherein, the map point of interest hierarchy relationship extraction model includes a hierarchy coefficient table of points of interest.

[0136] It should be noted that geographic entities occupy a specific physical location in the real world, possess relevant attributes, and exist independently. They can be natural geographic units or man-made facilities. Geographic entities are spatially related, and each geographic entity also has its own unique components. For example... Figure 3 As shown, the geographic entity full-space model is a conceptual model formed on the basis of the geographic entity concept, showing the relationships between various spatial entities and the relationships between various components. It is a full-space display of geographic entities based on the integration of above-ground and underground, indoor and outdoor spaces. The conceptual model is short for conceptual data model, a model of the real world for database users, mainly used to describe the conceptual structure of the world. Geographic entities include two categories: basic geographic entities and extended geographic entities. Basic geographic entities refer to entity objects that can be easily extracted and integrated from basic geographic information data results, including boundary entities, administrative region entities, road entities, railway entities, building entities, and courtyard entities; extended geographic entities are defined and integrated by geographic entity production and application departments according to specific data sources and application situations.

[0137] like Figure 3 As shown, geographic entity data consists of two layers: the primitive layer and the entity layer. Each type of entity corresponds to the spatial geometry of the entity, which is the primitive. Primitives are the basic units for spatial data expression and hierarchical organization, using point, line, and surface geometric elements. This is the same as the storage format of electronic map data, and is a map representation of the spatial location and geometric shape of physical features.

[0138] like Figure 4As shown, each graphic element is assigned a unique graphic element identifier, classification identifier, and lifecycle identifier, enabling information content classification and hierarchical management, as well as incremental updates based on graphic elements. The entity layer and geographic entities are composed of several graphic elements, each assigned a specific geographic meaning and possessing a unique geographic entity identifier. These identifiers are used to link geographic entities with relevant socio-economic and natural resource information. Geographic entities are typically categorized into point entities, line entities, area entities, and volume entities; complex geographic entities are composed of simpler entities. Different entities can be constructed based on the same graphic elements. Geographic entity data employs an entity-oriented modeling approach to describe independently existing "geographic entities" in the objective world.

[0139] Map data is often fragmented into multiple spatial targets when stored. Based on the connections in the real geographic world, there are corresponding connections between these targets. Currently, electronic map data is stored simply in the form of point, line, and polygon elements. A full-space model of geographic entities organizes these spatial targets through real-world connections, which can overcome the situation where geographic entities are fragmented due to the limitations of point, line, and polygon data representation.

[0140] It should also be noted that, such as Figure 5 As shown, the area involved in the hierarchical relationship of points of interest is the spatial relationship of building entities and the node connection relationship between building entities and road entities. Therefore, according to actual needs, building entity models are extracted from the pre-constructed geographic entity full-space model, points of interest are mapped to entity nodes, and the map point of interest hierarchical relationship extraction model is constructed according to the required hierarchical structure of points of interest.

[0141] The map interest point hierarchy relationship extraction model is established based on the pre-constructed full-space model logic of the geographic entities. Since interest points have similar attribute information to geographic entities, the elements in the interest point set can be mapped to these entity nodes, and all interest point elements are assigned corresponding hierarchy coefficients. That is, the hierarchy relationship between entity nodes is attached to the interest point elements, so that the interest point elements have a hierarchy relationship based on the building entity model in the full-space model of the geographic entities.

[0142] The hierarchy coefficients of the points of interest are shown in Table 1 below. The hierarchy coefficient of the central parent node of the courtyard level is 10; the hierarchy coefficient of the child nodes of the building level and courtyard type is 8; the hierarchy coefficient of the child nodes of the building level and building type is 6; and so on, the hierarchy coefficient of the child nodes of the floor level and building type is 4; finally, the hierarchy coefficient of the child nodes below the floor type is 2.

[0143]

[0144] Table 1

[0145] Optionally, such as Figure 6 As shown, step 103: Based on the pre-constructed map point of interest hierarchy relationship extraction model, the address semantic units corresponding to the address information of the points of interest, and the classification attributes corresponding to the name information of the points of interest, determine the hierarchy coefficients of the points of interest in the second set of points of interest, including:

[0146] Step 601: Map the points of interest to the map point of interest hierarchy relationship extraction model, and determine the hierarchy coefficient of the points of interest in the second set of points of interest as the first preset value;

[0147] It should be noted that, when mapping the points of interest to the map interest point hierarchy relationship extraction model, according to the hierarchy coefficient table shown in Table 1 above, the interest points in the fourth sub-interest point set P4 of the second interest point set are neither located within the scene plane nor within the building, have a simple spatial structure, simple attribute information, and no child nodes. Therefore, when mapped to other levels, the hierarchy coefficient of the interest points in the fourth sub-interest point set P4 is determined to be 2, and they are regarded as independent interest points.

[0148] The points of interest in the second sub-set of interest P2 within the second set of interest points can serve as parent nodes of courtyard levels or their child nodes. Determining the level of these points requires further judgment based on other attributes. Since these points have no spatial relationship with buildings, the hierarchical relationship of the points of interest in the second sub-set of interest P2 is determined only within the lower-level child nodes of the courtyard level; the hierarchical relationship is not passed down through building-type child nodes. The hierarchical coefficient of the points of interest in the second sub-set of interest P2 within the second set of interest points is determined to be 10 or 8, with the specific value requiring further determination. Particular attention should be paid to the point-like display elements on the scene surface, i.e., the points of interest in the third set of interest points N1 in step 202. These points of interest are the parent nodes of the courtyard level of all points of interest whose geographic coordinates are located within their corresponding scene surface. Their hierarchical level will be higher than all points of interest within that scene surface, and the hierarchical coefficient is determined to be 10.

[0149] The interest points in the third sub-interest point set P3 of the second interest point set are child points of the courtyard-level interest points, and their level is below that of the courtyard-level interest points. However, not all interest points can be the root nodes of the building level. For example, the difference between Company Z and Building A of Company Z. Therefore, interest points located in buildings still need to be gradually mapped to the map interest point level relationship extraction model according to their respective attribute information. The level coefficients of the interest points in the third sub-interest point set P3 are determined to be 8, 6, 4 and 2. The specific values ​​need to be further determined.

[0150] A portion of the interest points in the first sub-set P1 of the second set of interest points should serve as building-level parent nodes. The remaining interest points still need to be gradually mapped to the map interest point hierarchy extraction model based on their respective attribute information. The hierarchy coefficients of the interest points in the first sub-set P1 are 8, 6, 4, and 2, and the specific values ​​need to be further determined. Since they do not have any scene surfaces, there are no corresponding courtyard-level nodes. Not all interest point hierarchy relationships start from the courtyard level; interest point hierarchy relationships can start and end at any entity level.

[0151] Step 602: Based on the address semantic units corresponding to the address information of the points of interest, perform structured analysis to obtain the ending element, and determine the hierarchical coefficient of the points of interest in the second set of points of interest as the second preset value according to the ending element;

[0152] It should be noted that, as explained in step 301 above, the hierarchical coefficients of all interest points in the fourth sub-interest point set P4 and some interest points in the second sub-interest point set P2 have been determined. The other interest points in the second interest point set whose hierarchical coefficients have not been determined need to be further hierarchically divided. These interest points whose hierarchical coefficients have not been determined form the sixth interest point set N3, which includes all interest points in the first sub-interest point set P1, some interest points in the second sub-interest point set P2, and all interest points in the third interest point set P3, i.e., N3 = P1∪P′2∪P3;

[0153] Furthermore, the address information of the sixth set of points of interest (POIs) N3 is divided into multiple address semantic units. The spatial location of the POIs and the corresponding hierarchical coefficients in the map POI hierarchical relationship extraction model are extracted from the end elements. The rule-based address semantic unit extraction method relies on manually established rules, as shown in Table 2 below. The POI data processing method of this embodiment of the invention obtains the address information of the POIs based on big data analysis, which consists of the following parts:

[0154]

[0155] Table 2

[0156] Generally, each point of interest (POI) address information includes administrative division information, residential type, building number, and either a detailed address or location information. However, not every address strictly follows this one-to-one format; sometimes it may contain one or more of these information points, or it may lack them entirely. For example: "Building B, 22nd Floor, District D, City S," where "District D, City S" represents the administrative division, "Building B" represents the building information, and "22nd Floor" represents the detailed address. This address information lacks the street address, building number, and location information. Another example: "Building 8, No. 6, X Street, District H, City S, Province L," where "District H, City S, Province L" represents the administrative division, "No. 6, X Street" represents the street address, "Building 8" represents the building number, and "East" represents the location information. As shown in Table 3 below, using the POI elements of a certain prefecture-level city as a sample for analysis, a structured analysis of the address semantic units of all elements in the sample yielded 1230 address scenarios. Most of the ending elements are as follows:

[0157]

[0158] Table 3

[0159] As shown in Table 4, points of interest ending with a floor number can be directly labeled with the floor level attribute in the attribute table, with a level coefficient of 4. Points of interest ending with a door can be considered the lowest level point in the hierarchical relationship line of that point of interest. It may be a door of a building, i.e., a lower-level node of the building level, such as Door 2 of Dormitory 8 at Y University, or it may be a door of a courtyard, i.e., a lower-level node of the courtyard level, such as Door 1 of North Campus at Y University. Therefore, the level of points of interest ending with a door is classified as "Other", i.e., the level coefficient is 2. Points of interest ending with a room number / household name are similar to those ending with a door.

[0160] Points of interest ending with a street address typically include buildings or residential areas along the street. These points of interest must be at least at the building level. To determine the level of an interest, we need to consider the building number or residential type, the semantic unit above the address, and the name attribute information. For example, for "No. 47, P Road, Y Street, D District, S City," the level of interest cannot be determined solely based on the address information; it must be determined by combining it with P1, P2, and other relevant information. ′ Furthermore, interaction analysis of different spatial attributes in P3 reveals that points of interest located in P1 and P3 can have their hierarchical coefficients determined based on the address format. For example, in a residential complex, address number 56, East Wang First Street, Zone D, where building number 10 is address 56-6, East Wang First Street, Zone D. (The text abruptly shifts to a different topic) ′ Points of interest in a scene may be points of interest corresponding to that scene, such as the steel plant in area D, or points of interest contained within that scene, such as the outdoor activity area of ​​the steel plant in area D. Therefore, the hierarchy of such points of interest needs to be further determined based on other attribute information.

[0161]

[0162] Table 4

[0163] Points of interest ending with a village ID number are generally residences or factories located in suburban or rural areas. According to step 102, the set to which the point of interest belongs is determined, and its hierarchical coefficient is determined based on the semantic units in the name information. Since the concept of a village ID number already limits the scope of the point of interest to areas outside urban areas such as suburbs or towns, and through interaction analysis with the different spatial attributes of P1, P′2, and P3, it can be seen that: when the point of interest belongs to the third sub-point of interest set P3 and the ending element is a village ID number, since the village ID number is a relatively independent building unit number within a township, this type of point of interest is located at the building level; when the point of interest belongs to the second sub-point of interest set P2 and the ending element is a village ID number, for the same reason, this type of point of interest is located at the courtyard level; when the point of interest belongs to the first sub-point of interest set P1 and the ending element is a village ID number, this type of point of interest is located at the building level.

[0164] Step 603: When the ending element includes a road address or directional word, the name information of the point of interest is segmented using a word segmentation engine to determine the semantic unit and semantic unit combination method corresponding to the name information, and the hierarchical coefficient of the point of interest in the second set of points of interest is determined as a third preset value based on the classification attribute of the semantic unit and the semantic unit combination method.

[0165] It should be noted that after structured analysis based on the address semantic units corresponding to the address information of points of interest, the hierarchical coefficients of some points of interest were determined. However, the hierarchical coefficients of some points of interest remain undetermined. For example, for points of interest whose address information ends with a directional word, it is necessary to determine the information of the semantic unit at the next higher level than the ending element, and simultaneously extract the classification attribute corresponding to the name information of the point of interest to obtain its hierarchical coefficient. Since directional words contain a large amount of information and lack obvious classification information, points of interest whose ending elements include directional words need to have their hierarchical coefficients determined through other methods. In addition, there are still some points of interest whose address information does not contain the aforementioned ending elements. All points of interest with undetermined hierarchical coefficients form the seventh set of points of interest, N4. Next, the name information of the points of interest in this set will be extracted to further determine the hierarchical coefficients.

[0166] The names of points of interest (POIs) exist in text form, allowing users to focus on them intuitively. As a detailed description of the POIs, they possess certain categorization attributes. The POI names are segmented into multiple semantic units, and their uses are differentiated based on similarity. Furthermore, by identifying the usage of proper nouns, categorization attributes are extracted from the POI names to determine their hierarchical structure.

[0167] The structure of a point of interest's name information is a combination of modifiers and headwords. Words in different positions have different meanings and play different roles. For example, in "B University Second Cafeteria," "cafeteria" is the headword of this point of interest's name information. "B University" and "second" are both modifiers of "cafeteria," but "university" in "B University" is the headword of this modifier, used to emphasize that this "cafeteria" is a cafeteria within a "university." During the word segmentation process of a point of interest's name information, more detailed category information can be extracted to determine the hierarchy.

[0168] Furthermore, the name information of the points of interest is segmented using a word segmentation engine to obtain the smallest possible semantic unit combination: W = w1w2w3…w m ;

[0169] For a part-of-speech matrix Q with multiple meanings or contexts:

[0170]

[0171] And the optimal combination of semantic units is determined by using a hidden Markov model:

[0172] Q i =argmaxP(M|W);

[0173] W, as a combination of POI names, is the set of all observations, i.e., the observation sequence; M, as the set of all state values, is the state sequence.

[0174] Furthermore, based on the Hidden Markov Model, the optimal semantic unit combination for the name information of the seventh set of interest points N4 is obtained. Based on this, semantic units of interest related to classification attributes are extracted from the name information of the seventh set of interest points N4. For interest points whose ending elements include road address numbers and are located only within the scene plane, it is observed whether their central word is a unit containing courtyard information. If the central word of the interest point's name information falls on a semantic unit containing courtyard information, then the interest point's level is courtyard; otherwise, the interest point is a child node of the courtyard level. Since it is not located within the building plane, the interest point is only a child node of the courtyard level, not a parent node of the building level.

[0175] For points of interest (POIs) whose ending element includes a directional word and are located within a building, determine whether their central word is an independent building type. If the central word of the POI's name information is an independent building type noun, then the POI's level is "building". If the central word of the POI's name information is a building type noun, such as "xx block xx building xx unit...", then the POI's level is "building". If the central word is not a specific building or building type noun, then the POI's level is a bottom-level POI. For POIs whose ending element includes a directional word and are only located within the scene plane, observe whether their central word is a unit containing courtyard information. If the central word of the POI's name information falls on a semantic unit of courtyard information, then the POI's level is "courtyard". Otherwise, the POI is a child node of the courtyard level. Since it is not located within the building roof, the POI is only a child node of the courtyard level, not a parent node of the building level. For other POIs that did not participate in the structured analysis of address information, extract keywords such as "courtyard", "building", "building", and "floor" from the semantic units corresponding to the POI's name information to determine the level.

[0176] It should also be noted that the point of interest (POI) data processing method described in this embodiment of the invention establishes connections between map data of different formats to supplement and enrich the POI set to be processed, making full use of the spatial relationships between map data and improving data utilization. It uses a combination of semantic similarity and character similarity to identify entities with the same name, and through accuracy and quality assessment, achieves data deduplication, improving the accuracy of POI data. Based on the building entity model in the geographic entity full-space model, a map POI hierarchical relationship extraction model is constructed, mapping POIs to this model and determining the hierarchical coefficients of POIs for subsequent cartographic work, making the electronic map display more reasonable and improving the utilization rate of the electronic map.

[0177] like Figure 7 As shown, this embodiment of the invention also provides a point-of-interest (POI) data processing device, comprising:

[0178] The processing module 701 is used to supplement the set of interest points to be processed using scene surface data and perform deduplication to obtain the first set of interest points;

[0179] The partitioning module 702 is used to partition the spatial types of the interest points in the first interest point set according to the spatial relationships of the interest points, so as to obtain at least one second interest point set.

[0180] The determination module 703 is used to determine the hierarchical coefficients of the interest points in the second set of interest points based on the pre-built map interest point hierarchy relationship extraction model, the address semantic units corresponding to the address information of the interest points, and the classification attributes corresponding to the name information of the interest points.

[0181] In this embodiment of the invention, a first set of points of interest (POIs) is obtained by supplementing the set of POIs to be processed with scene surface data and performing deduplication. Based on the spatial relationships of the POIs, the spatial types of the POIs in the first set of POIs are divided to obtain at least one second set of POIs. Based on a pre-constructed map POI hierarchy relationship extraction model, the address semantic units corresponding to the address information of the POIs, and the classification attributes corresponding to the name information of the POIs, the hierarchy coefficients of the POIs in the second set of POIs are determined. This fully utilizes the spatial relationships between map data, improves data utilization, and clarifies the hierarchy coefficients of the POIs, facilitating electronic map production, thereby improving the display effect of electronic maps and increasing their utilization rate.

[0182] Optionally, such as Figure 8 As shown, the processing module 701 includes:

[0183] The first processing unit 801 is used to extract spatial information, category information and attribute information from the scene surface data, and obtain the set of interest points corresponding to the scene surface data through dimensionality reduction processing.

[0184] The supplementary unit 802 is used to supplement the set of interest points to be processed according to the set of interest points corresponding to the scene surface data, so as to obtain a third set of interest points;

[0185] Extraction unit 803 is used to extract interest points of entities with the same name in the third interest point set to obtain a fourth interest point set;

[0186] The second processing unit 804 is used to perform deduplication processing on the fourth set of interest points according to preset rules to obtain a fifth set of interest points to be deleted.

[0187] The deletion unit 805 is used to delete the points of interest in the third set of points of interest that correspond to the fifth set of points of interest, so as to obtain the first set of points of interest.

[0188] Optionally, the extraction unit 803 is specifically used for:

[0189] Using the spatial, category, and attribute information of the interest point set corresponding to the scene surface data, the third interest point set is searched for category attribute similarity and spatial similarity, and interest points of entities with the same name in the third interest point set are extracted to obtain the fourth interest point set.

[0190] Among them, at least two points of interest whose name and address information meet at least one of the following criteria are considered to be points of interest of the same entity:

[0191] The name information is the same and the address information is the same;

[0192] The name and address information meet the first similarity condition;

[0193] The name and address information meet the second similarity condition;

[0194] The name and address information meet the third similarity condition.

[0195] Optionally, the preset rules include at least one of the following:

[0196] If at least two points of interest in the fourth set of points of interest have the same name information and the same address information, or if the name information and address information meet the first similarity condition, one point of interest is retained.

[0197] When the name and address information of at least two interest points in the fourth interest point set meet the second similarity condition, the interest point with the correct positional relationship with the scene plane is retained.

[0198] When the name and address information of at least two points of interest in the fourth set of points of interest meet the third similarity condition, the point of interest with the higher accuracy quality score is retained.

[0199] Optionally, the device further includes:

[0200] The acquisition module is used to obtain the ratio of the number of semantic units in the name of a point of interest to the distance from that point of interest to the corresponding scene surface;

[0201] The module is used to obtain the accuracy quality score based on the ratio.

[0202] Optionally, the at least one set of second points of interest includes at least one of the following:

[0203] The point of interest is the first set of sub-points of interest located within the building;

[0204] The set of second sub-interest points located within the scene plane but not within buildings;

[0205] The set of third sub-interest points located within the scene plane and within buildings;

[0206] The fourth set of sub-interest points that are neither located within the scene plane nor within a building.

[0207] Optionally, the device further includes:

[0208] The construction module is used to map points of interest to entity nodes based on the building entity models in the pre-built geographic entity full-space model, and to construct the map point of interest hierarchy relationship extraction model; wherein, the map point of interest hierarchy relationship extraction model includes a hierarchy coefficient table of points of interest.

[0209] Optionally, the determining module 703 is specifically used for:

[0210] The points of interest are mapped to the map point of interest hierarchy extraction model, and the hierarchy coefficient of the points of interest in the second set of points of interest is determined as a first preset value;

[0211] Based on the address semantic units corresponding to the address information of the points of interest, a structured analysis is performed to obtain the ending element, and based on the ending element, the hierarchical coefficient of the points of interest in the second set of points of interest is determined as the second preset value.

[0212] When the ending element includes a road address or directional word, the name information of the point of interest is segmented using a word segmentation engine to determine the semantic unit and semantic unit combination method corresponding to the name information. Based on the classification attribute of the semantic unit and the semantic unit combination method, the hierarchical coefficient of the point of interest in the second set of points of interest is determined to be a third preset value.

[0213] It should be noted that the point of interest data processing device provided in the embodiments of the present invention is a device capable of executing the above-described point of interest data processing method. Therefore, all embodiments of the above-described point of interest data processing method are applicable to this device and can achieve the same or similar technical effects.

[0214] like Figure 9 As shown, this embodiment of the invention also provides an electronic device, including: a processor 901, a memory 902, and a program stored in the memory 902 and executable on the processor 901. When the program is executed by the processor 901, it implements the steps of the point of interest data processing method as described above.

[0215] Optionally, the electronic device further includes a transceiver 903 for receiving and sending data under the control of the processor 901.

[0216] Among them, Figure 9In this context, the bus architecture can include any number of interconnected buses and bridges, specifically linking various circuits together, represented by one or more processors (processor 901) and memory (memory 902). The bus architecture can also link various other circuits, such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. The bus interface provides a user interface 904. The transceiver 903 can be multiple elements, including transmitters and receivers, providing a unit for communicating with various other devices over a transmission medium. For different electronic devices, the user interface 904 can also be an interface capable of connecting external or internal devices, including but not limited to keypads, displays, speakers, microphones, joysticks, etc. The processor 901 is responsible for managing the bus architecture and general processing, and the memory 902 can store data used by the processor 901 during operation.

[0217] This invention also provides a readable storage medium storing a program that, when executed by a processor, implements the steps of the point of interest data processing method described above.

[0218] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0219] It should be further noted that the electronic devices described in this specification include, but are not limited to, smartphones, tablets, etc., and many of the described functional components are referred to as modules in order to more specifically emphasize the independence of their implementation.

[0220] In this embodiment of the invention, the module can be implemented in software so that it can be executed by various types of processors. For example, an identified executable code module may include one or more physical or logical blocks of computer instructions, which may be constructed as objects, procedures, or functions. Nevertheless, the executable code of the identified module does not need to be physically located together, but may include different instructions stored in different bits, which, when logically combined, constitute the module and achieve the module's intended purpose.

[0221] In practice, an executable code module can be a single instruction or many instructions, and can even be distributed across multiple different code segments, different programs, and across multiple memory devices. Similarly, operational data can be identified within the module and can be implemented in any suitable form and organized within any suitable type of data structure. This operational data can be collected as a single dataset or distributed across different locations (including different storage devices), and can exist, at least in part, solely as electronic signals within the system or network.

[0222] When a module can be implemented using software, considering the current level of hardware technology, modules that can be implemented in software can be implemented using hardware circuits by those skilled in the art to achieve the corresponding functions, without considering cost. These hardware circuits include conventional very-large-scale integrated circuits (VLSI) or gate arrays, as well as existing semiconductors such as logic chips and transistors, or other discrete components. Modules can also be implemented using programmable hardware devices, such as field-programmable gate arrays, programmable array logic, and programmable logic devices.

[0223] The exemplary embodiments described above are with reference to the accompanying drawings. Many different forms and embodiments are feasible without departing from the spirit and teachings of the invention. Therefore, the invention should not be construed as limiting the exemplary embodiments set forth herein. Rather, these exemplary embodiments are provided to make the invention complete and convey the scope of the invention to those skilled in the art. In these drawings, component dimensions and relative dimensions may be exaggerated for clarity. The terminology used herein is for the purpose of describing particular exemplary embodiments only and is not intended to be limiting. As used herein, unless clearly indicated otherwise, the singular forms “a,” “an,” and “the” are intended to include all such forms. It will be further understood that the terms “comprising” and / or “including”, when used in this specification, indicate the presence of the stated features, integers, steps, operations, components, and / or elements, but do not exclude the presence or addition of one or more other features, integers, steps, operations, components, and / or groups thereof. Unless otherwise indicated, when stated, a range of values ​​includes the upper and lower limits of the range and any subranges in between.

[0224] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for processing point-of-interest (POI) data, characterized in that, include: The set of interest points to be processed is supplemented by scene surface data and deduplication is performed to obtain the first set of interest points; Based on the spatial relationships of the points of interest, the spatial types of the points of interest in the first set of points of interest are divided to obtain at least one second set of points of interest; the spatial relationships of the points of interest include the spatial relationships between the points of interest and the scene plane, and the spatial relationships between the points of interest and the buildings. Based on a pre-constructed map point of interest hierarchy extraction model, address semantic units corresponding to the address information of the points of interest, and classification attributes corresponding to the name information of the points of interest, the hierarchy coefficients of the points of interest in the second set of points of interest are determined; the map point of interest hierarchy extraction model is constructed by mapping the points of interest to entity nodes based on the building entity model in the pre-constructed geographic entity full-space model, and the map point of interest hierarchy extraction model includes a hierarchy coefficient table of points of interest. Specifically, based on a pre-constructed map point of interest hierarchy extraction model, address semantic units corresponding to the address information of the points of interest, and classification attributes corresponding to the name information of the points of interest, the hierarchy coefficients of the points of interest in the second set of points of interest are determined, including: The points of interest are mapped to the map point of interest hierarchy extraction model, and the hierarchy coefficient of the points of interest in the second set of points of interest is determined as a first preset value; Based on the address semantic units corresponding to the address information of the points of interest, a structured analysis is performed to obtain the ending element, and based on the ending element, the hierarchical coefficient of the points of interest in the second set of points of interest is determined as the second preset value. When the ending element includes a road address or directional word, the name information of the point of interest is segmented using a word segmentation engine to determine the semantic unit and semantic unit combination method corresponding to the name information. Based on the classification attribute of the semantic unit and the semantic unit combination method, the hierarchical coefficient of the point of interest in the second set of points of interest is determined to be a third preset value.

2. The point-of-interest data processing method according to claim 1, characterized in that, The set of interest points to be processed is supplemented using scene surface data, and deduplication is performed to obtain the first set of interest points, including: Spatial information, category information, and attribute information are extracted from the scene surface data, and the set of interest points corresponding to the scene surface data is obtained through dimensionality reduction processing. Based on the set of interest points corresponding to the scene surface data, the set of interest points to be processed is supplemented to obtain a third set of interest points; Extract the interest points of entities with the same name from the third set of interest points to obtain the fourth set of interest points; According to preset rules, the fourth set of interest points is deduplicated to obtain the fifth set of interest points to be deleted. The interest points corresponding to the fifth interest point set in the third interest point set are deleted to obtain the first interest point set.

3. The point-of-interest data processing method according to claim 2, characterized in that, Extracting the interest points of entities with the same name from the third set of interest points yields a fourth set of interest points, including: Using the spatial, category, and attribute information of the interest point set corresponding to the scene surface data, the third interest point set is searched for category attribute similarity and spatial similarity, and interest points of entities with the same name in the third interest point set are extracted to obtain the fourth interest point set. Among them, at least two points of interest whose name and address information meet at least one of the following criteria are considered to be points of interest of the same entity: The name information is the same and the address information is the same; The name and address information meet the first similarity condition; The name and address information meet the second similarity condition; The name and address information meet the third similarity condition.

4. The point-of-interest data processing method according to claim 2, characterized in that, The preset rules include at least one of the following: If at least two points of interest in the fourth set of points of interest have the same name information and the same address information, or if the name information and address information meet the first similarity condition, one point of interest is retained. When the name and address information of at least two interest points in the fourth interest point set meet the second similarity condition, the interest point with the correct positional relationship with the scene plane is retained. When the name and address information of at least two points of interest in the fourth set of points of interest meet the third similarity condition, the point of interest with the higher accuracy quality score is retained.

5. The point-of-interest data processing method according to claim 4, characterized in that, Also includes: Get the ratio of the number of semantic units in the name of a point of interest to the distance from that point of interest to the corresponding scene surface; The accuracy quality score is obtained based on the ratio.

6. The method for processing point-of-interest data according to claim 1, characterized in that, The at least one set of second points of interest includes at least one of the following: The point of interest is the first set of sub-points of interest located within the building; The set of second sub-interest points located within the scene plane but not within buildings; The set of third sub-interest points located within the scene plane and within buildings; The fourth set of sub-interest points that are neither located within the scene plane nor within a building.

7. An interest point data processing device, characterized in that, include: The processing module is used to supplement the set of interest points to be processed using scene surface data and perform deduplication to obtain the first set of interest points; The segmentation module is used to segment the spatial types of the interest points in the first set of interest points according to the spatial relationships of the interest points, so as to obtain at least one second set of interest points; the spatial relationships of the interest points include the spatial relationships between the interest points and the scene plane, and the spatial relationships between the interest points and the buildings. The determination module is used to determine the hierarchy coefficients of the points of interest in the second set of points of interest based on a pre-built map point of interest hierarchy relationship extraction model, the address semantic units corresponding to the address information of the points of interest, and the classification attributes corresponding to the name information of the points of interest. The map point of interest hierarchy relationship extraction model is constructed by mapping the points of interest to entity nodes based on the building entity model in the pre-built geographic entity full-space model, and the map point of interest hierarchy relationship extraction model includes a hierarchy coefficient table of points of interest. The determining module is specifically used for: The points of interest are mapped to the map point of interest hierarchy extraction model, and the hierarchy coefficient of the points of interest in the second set of points of interest is determined as a first preset value; Based on the address semantic units corresponding to the address information of the points of interest, a structured analysis is performed to obtain the ending element, and based on the ending element, the hierarchical coefficient of the points of interest in the second set of points of interest is determined as the second preset value. When the ending element includes a road address or directional word, the name information of the point of interest is segmented using a word segmentation engine to determine the semantic unit and semantic unit combination method corresponding to the name information. Based on the classification attribute of the semantic unit and the semantic unit combination method, the hierarchical coefficient of the point of interest in the second set of points of interest is determined to be a third preset value.

8. An electronic device, characterized in that, include: A processor, a memory, and a program stored in the memory and executable on the processor, wherein the program, when executed by the processor, implements the steps of the point of interest data processing method as described in any one of claims 1 to 6.

9. A readable storage medium, characterized in that, The readable storage medium stores a program that, when executed by a processor, implements the steps of the point of interest data processing method as described in any one of claims 1 to 6.

Images