A city functional area identification method based on POI and improved topic model

Abstract

The application discloses a city functional area identification method based on POI and an improved topic model, and belongs to the technical field of geographic information systems. The city functional area identification method based on POI and the improved topic model comprises the following steps: obtaining interest point data of a target functional area, wherein the interest point data comprises spatial position data of each interest point; dividing the target functional area into a plurality of functional sub-areas according to the spatial position data; determining sub-area semantic features of each functional sub-area; and obtaining regional spatial semantic features of the target functional area according to the sub-area semantic features, so that the problems of low recognition precision and poor accuracy existing in the prior art are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geographic information systems, and in particular to a city functional area identification method based on POI and an improved topic model. BACKGROUND

[0002] With the advent of the big data era, scholars have gradually carried out in-depth research on the identification of city functional areas. The identification of city functional areas based on POI is one of the important directions. This is because POI (Points of Interest) data has the characteristics of availability and integrity, and more importantly, POI not only contains the physical characteristics of geographic entities, but also represents the social and economic attribute characteristics, which is in line with the characteristics of city functional areas representing both regional physical characteristics and social and economic characteristics.

[0003] At present, the method based on the topic model in the city functional area identification method based on POI adopts the bag-of-words model to represent the POI data, takes the city functional area as a document, and takes the POI in the functional area as a word of the document. According to the frequency of the POI in the functional area, the semantic information in the functional area is mined from the co-occurrence relationship of the POI, and the city functional area is identified according to the semantic information.

[0004] The topic model is a common method in natural language processing. However, unlike the words in natural language, the POI data that constitutes the city functional area has spatial characteristics, resulting in spatial heterogeneity of the semantic characteristics of the functional area. The existing research uses the topic model to extract the semantic information of the city functional area, ignores the spatial characteristics of the POI and the spatial heterogeneity of the semantic characteristics of the functional area, and restricts the accuracy of the expression of the semantic characteristics, thereby causing the problems of low identification accuracy and poor accuracy of the functional area.

[0005] CONTENT

[0006] The main purpose of the present application is to provide a city functional area identification method based on POI and an improved topic model, which aims to solve the problems of low identification accuracy and poor accuracy in the existing functional area identification technology.

[0007] To achieve the above purpose, the present application provides a city functional area identification method based on POI and an improved topic model, which comprises:

[0008] acquiring point of interest data of a target functional area, wherein the point of interest data comprises spatial position data of each point of interest;

[0009] dividing the target functional area into a plurality of functional sub-areas according to the spatial position data;

[0010] determine sub-region semantic features of each of the functional sub-regions;

[0011] obtain a region spatial semantic feature of the target functional region according to the sub-region semantic features.

[0012] Optionally, the interest point data of the target functional region is obtained, wherein the interest point data comprises spatial position data of each interest point, and the spatial position data comprises:

[0013] The city interest point data of a city where the target functional region is located is obtained, wherein the interest point data comprises spatial position data of interest points.

[0014] The city interest point data is matched to a city map according to the spatial position data of the interest points, to obtain interest point data of the target functional region.

[0015] Optionally, the target functional region is divided into a plurality of functional sub-regions according to the spatial position data, and the method comprises:

[0016] The interest points are clustered according to the spatial position data, to obtain a first clustering result.

[0017] The target functional region is divided into a plurality of sub-regions according to the first clustering result.

[0018] Optionally, the sub-region semantic features of each of the functional sub-regions are determined, and the method comprises:

[0019] A topic distribution of each of the functional sub-regions is determined according to an LDA model and all interest point data of the functional sub-regions.

[0020] The topics with a probability arranged in a front preset number of topics in the identified topic distribution are determined as the sub-region semantic features of the functional sub-regions.

[0021] Optionally, the region spatial semantic feature of the target functional region is obtained according to the sub-region semantic features, and the method comprises:

[0022] The sub-region semantic features of all the functional sub-regions are clustered, to obtain a feature type of the target functional region.

[0023] Optionally, if the sub-region semantic features of each of the functional sub-regions belong to the same type, the sub-region semantic features are taken as the region spatial semantic feature of the target functional region.

[0024] Optionally, if the sub-region semantic features of each of the functional sub-regions do not belong to the same type, the sub-region semantic features of all the functional sub-regions are clustered, to obtain a second clustering result.

[0025] According to the second clustering result, a region spatial semantic feature of the target functional area is determined.

[0026] In addition, to achieve the above object, the application further provides a device for identifying a city functional area based on POI and an improved topic model, which comprises:

[0027] a data acquisition module, configured to acquire point of interest (POI) data of a target functional area, wherein the POI data comprises spatial position data of each POI;

[0028] a region division module, configured to divide the target functional area into a plurality of sub-regions according to the spatial position data.

[0029] a type determination module, configured to determine a sub-region semantic feature of each functional sub-region;

[0030] a type identification module, configured to obtain a region spatial semantic feature of the target functional area according to the sub-region semantic feature.

[0031] The application further provides a device for identifying a city functional area based on POI and an improved topic model, which comprises a processor, a memory and a program for identifying a city functional area based on POI and an improved topic model stored in the memory, wherein the program for identifying a city functional area based on POI and an improved topic model, when executed by the processor, implements each step in any possible implementation manner of the implementation manners of the first aspect.

[0032] The application further provides a computer readable storage medium, which stores a program for identifying a city functional area based on POI and an improved topic model, wherein the program for identifying a city functional area based on POI and an improved topic model, when executed by a processor, implements each step in any possible implementation manner of the implementation manners of the first aspect.

[0033] The embodiment of the present application proposes a city functional area identification method based on POI and improved topic model. The spatial position data of the interest points is obtained by acquiring the interest point data of the target functional area. The target functional area is divided into multiple functional sub-areas according to the spatial position data of the interest points. The sub-area semantic features in each functional sub-area are determined. The semantic features of the target functional area are further obtained by fusing the semantic features of the sub-areas in the target functional area. That is, the present application: (1) first, based on the spatial position information of the interest points, the spatial clustering algorithm is used to cluster the interest points, and the spatially adjacent POIs are divided into a sub-area of the target functional area and combined as a sub-area of the target functional area. Therefore, the interest points in the same sub-functional area have stronger spatial correlation. (2) The semantic features of the sub-area are extracted by using the topic model. In this step, the sub-area is taken as the spatial scale of extracting the semantic features of the target functional area, which is different from directly applying the topic model to extract the semantic features of the functional area. Based on the interest points with stronger spatial correlation, the semantic features of each functional sub-area can be more accurately identified, and the spatial correlation of the interest points in the same semantic feature is ensured. (3) The semantic features of the sub-areas in the functional area are fused to further obtain the semantic features of the target functional area. The feature contains the spatial characteristics of the semantic features, and reflects the spatial heterogeneity of the semantic features in the functional area.

[0034] The present application redefines the spatial scale of extracting the semantic features (topics) of the target functional area by dividing the sub-areas in the target functional area. The spatial scale constraint of semantic feature extraction realizes the extraction of the spatial semantic features of the functional area. In addition, the present application can avoid the identification error caused by the POIs in the high-frequency part of the functional sub-area covering the POIs in the remaining part of the functional sub-area which are highly aggregated but have low appearance frequency when directly using the topic model to identify the functional area. Therefore, the accuracy and precision of the functional area identification are improved. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 The structure diagram of the city functional area identification device based on POI and improved topic model of the hardware running environment involved in the present application;

[0036] Figure 2 The flowchart of the first embodiment of the city functional area identification method based on POI and improved topic model of the present application;

[0037] Figure 3 The flowchart of the second embodiment of the city functional area identification method based on POI and improved topic model of the present application;

[0038] Figure 4A distribution diagram of functional area 1 and POIs in functional area 1;

[0039] Figure 5 A distribution diagram of functional area 2 and POIs in functional area 2;

[0040] Figure 6 A functional module diagram of the device for identifying a city functional area based on POIs and improved topic models according to the present application.

[0041] The implementation, functional characteristics and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0042] It should be understood that the specific embodiments described herein are merely intended to explain the present application and not to limit the present application.

[0043] Since the prior art method based on a word embedding model, the POI type-based features are adjacent to identify the city functional area, and the POI type is an object-oriented and low-level feature, which cannot express higher-level semantic information; the method based on a topic model uses a bag-of-words model to represent the POI data, and the bag-of-words model only considers the frequency of the POIs in the functional area, and ignores the spatial features of the POIs, resulting in a reduction in the recognition accuracy. Therefore, the above two methods for identifying the city functional area cannot realize the identification of the city functional area based on the spatial semantics of the POIs, so as to improve the accuracy and accuracy of the city functional area identification.

[0044] To this end, the application provides a city functional area identification method based on POI and improved topic model, spatial position data of interest points is obtained by acquiring interest point data of a target functional area; the target functional area is divided into multiple functional sub-areas according to the spatial position relationship of the interest points; the semantic features of the sub-areas in each functional sub-area are determined; and the semantic features of the sub-areas in the target functional area are fused to further obtain the semantic features of the target functional area. That is, the application: (1) first, based on the spatial position information of the interest points, a clustering algorithm is used to spatially cluster the interest points, and POIs that are adjacent in space are divided together to combine a sub-area of the target functional area, so that the interest points in the same sub-functional area have stronger spatial correlation. (2) The semantic features of the sub-areas are extracted using a topic model. In this step, the method takes the sub-area as the spatial scale for extracting the semantic features of the target functional area, which is different from directly applying the topic model to extract the semantic features of the functional area, and can more accurately identify the semantic features of each functional sub-area based on the interest points with stronger spatial correlation, thereby ensuring the spatial correlation of the interest points in the same semantic feature. (3) The semantic features of the sub-areas in the functional area are fused to further obtain the semantic features of the target functional area, and the features contain the spatial characteristics of the semantic features and reflect the spatial heterogeneity of the semantic features in the functional area.

[0045] The application redefines the spatial scale of extracting the semantic features (topics) of the target functional area by dividing the sub-areas in the target functional area, and realizes the extraction of the spatial semantic features of the functional area by the spatial scale constraint of the semantic feature extraction. In addition, the application can also avoid the identification error caused by the POIs in part of the functional sub-area that are high-frequency and mask the POIs in the remaining part of the functional sub-area that are highly aggregated but have low frequency, when directly using the topic model to identify the functional area, thereby improving the precision and accuracy of the functional area identification.

[0046] As Figure 1As shown, the city functional area identification device based on POI and improved topic model can include a processor 1001, for example, a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to realize the connection and communication between the components. The user interface 1003 can include a display, an input unit such as a keyboard, and can also include a standard wired interface, a wireless interface. The network interface 1004 can optionally include a standard wired interface, a wireless interface (such as a wireless fidelity (WIreless-FIdelity, WI-FI) interface). The memory 1005 can be a high-speed random access memory (RAM) memory, or a stable non-volatile memory (NVM) such as a disk memory. The memory 1005 can also be a storage device independent of the aforementioned processor 1001.

[0047] Those skilled in the art can understand that, Figure 1 The structure shown in the figure does not constitute a limitation on the terminal, and can include more or fewer components than the figure, or combine certain components, or different component arrangements.

[0048] As Figure 1 As shown, the memory 1005 as a storage medium can include an operating system, a data storage module, a network communication module, a user interface module, and a city functional area identification program based on POI and improved topic model.

[0049] As Figure 1 As shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the server of the present application can be arranged in the server, the server calls the city functional area identification program based on POI and improved topic model stored in the memory 1005 through the processor 1001, and executes the city functional area identification method based on POI and improved topic model provided by the embodiment of the present application.

[0050] Based on the above hardware structure but not limited to the above hardware structure, the present application provides a first embodiment of a city functional area identification method based on POI and improved topic model. Referring to Figure 2 , Figure 2 The first embodiment of the city functional area identification method based on POI and improved topic model of the present application is a flowchart.

[0051] Referring to Figure 2 The first embodiment of the present application provides a city functional area identification method based on POI and improved topic model, the method comprising:

[0052] S10: obtaining interest point data of a target functional area, wherein the interest point data comprises spatial position data of each interest point;

[0053] It should be noted that the method execution subject of the embodiment is a city functional area identification system, which is used to obtain and identify interest point data of a target functional area, and can be a local calculator or a cloud server.

[0054] The target functional area can be a region in a city, and the area size can be set according to actual needs.

[0055] The interest point POI refers to a point on an electronic map representing a geographical entity, which contains physical features and socio-economic attribute features of the geographical entity. The POI not only contains the physical features of the geographical entity, but also represents the socio-economic attribute features. The spatial position data is the latitude and longitude coordinate data of the interest point on the electronic map. The above interest point data further comprises a third-level type of the interest point.

[0056] The city functional area identification system obtains data of interest points on an electronic map in a target functional area of a city, for example, the city functional area identification system obtains latitude and longitude coordinate data and third-level type data of entity points on the electronic map of a region B in an A city as interest point data of the region B.

[0057] S20: dividing the target functional area into a plurality of functional sub-regions according to the spatial position data;

[0058] The functional sub-region is obtained by dividing the target functional area by the city functional area identification system and the aggregation degree of the interest points in the target functional area. In this step, spatially adjacent POIs are divided together to form a sub-region, so that the interest points in the same sub-functional region have stronger spatial correlation.

[0059] For example, all interest points in the region B are densely distributed at the upper left corner, uniformly distributed at the upper right corner, and sparsely distributed at the lower left and right corners on the electronic map, and then the region B is divided into a sub-region b1 (upper left corner), a sub-region b2 (upper right corner), and a sub-region b3 (lower).

[0060] S30: determining a sub-region semantic feature of each functional sub-region;

[0061] The sub-region semantic feature is the POI type with the highest frequency among the POI types contained in the identified functional sub-region.

[0062] The urban functional area recognition system obtains the sub-area semantic feature of each functional sub-area according to the third type of data of the interest points in each functional sub-area. This step takes the sub-area as the target function area to extract the semantic feature, and can more accurately identify the sub-area semantic feature of each functional sub-area based on the interest points with stronger spatial correlation, thereby ensuring the spatial correlation of the interest points in the same semantic feature.

[0063] S40: Obtain the area spatial semantic feature of the target functional area according to the sub-area semantic feature;

[0064] The area spatial semantic feature is the regional sub-area semantic feature of the target functional area, and can be a residential area, a commercial area, a living area, etc. It can be understood that the area spatial semantic feature can include one or more. The urban functional area recognition system obtains the area spatial semantic feature of the target functional area further based on the sub-area semantic features of the functional sub-areas. This step integrates the semantic features of the sub-areas in the functional area to further obtain the semantic feature of the target functional area, which contains the spatial characteristics of the semantic feature and reflects the spatial heterogeneity of the semantic feature in the functional area.

[0065] It can be easily understood that in the prior art, the LDA model is affected by the number of POI classes in the area when extracting the area spatial semantic feature. In other words, the result of the LDA model ignores the influence of the relatively small number of POI classes on the area function, and only shows the area function determined by the relatively large number of POI classes. For example, for a composite functional area of a residential area and a commercial area, generally, the number of commercial POIs is significantly larger than the number of residential POIs, which leads to the error identification of the composite functional area of the residential area and the commercial area as a commercial area when the LDA model is used to identify the area spatial semantic feature.

[0066] In this embodiment, the target functional area is first divided into multiple sub-areas according to the spatial positions of the interest points, so that the POI data contains spatial features. Therefore, the closely related POI classes are divided together, the correlation of the interest points in the same sub-functional area is stronger, and the functions of these sub-functional areas are more single. When the method based on the topic model is used to identify the sub-area semantic feature of the sub-functional area, the sub-area semantic feature of each sub-functional area can be more accurately distinguished without being affected by the high-frequency POI classes of other functional sub-areas. Then, the area spatial semantic feature of the target functional area as a whole is obtained based on the sub-area semantic features of the sub-functional areas. Thus, the target functional area is identified based on the spatial semantic feature of the interest points in this embodiment, thereby improving the accuracy and precision of the functional area recognition.

[0067] Referring to Figure 3 , Figure 3 A flowchart of a second embodiment of a city functional area identification method based on POI and an improved topic model is provided for the application, based on the above Figure 2 The second embodiment of the city functional area identification method based on POI and the improved topic model is provided for the application, based on the above

[0068] In this embodiment, the method comprises:

[0069] S11: Obtain city interest point data of a target functional area, wherein the interest point data comprises spatial position data of the interest point;

[0070] S12: Match the city interest point data to a city map according to the spatial position data of the interest point, to obtain interest point data of the target functional area;

[0071] The city map is an electronic map, for example, the city can be displayed by using Gaode Map APP.

[0072] The city functional area identification system obtains the data of the interest point, including the latitude and longitude coordinates of the interest point and the third-level type, and matches the interest point to the electronic map of the city according to the obtained latitude and longitude coordinates of the interest point, to obtain the interest point data of the target functional area.

[0073] S21: Cluster the interest points according to the spatial position data, to obtain a first clustering result;

[0074] S22: Divide the target functional area into multiple sub-areas according to the first clustering result;

[0075] The clustering mode can be a spatial clustering mode of automatically determining the number of categories, which refers to dividing the objects in a spatial data set into classes composed of similar objects, the objects in the same class have high similarity, and the objects in different classes have low similarity, and finally each class is recorded as a functional sub-area.

[0076] The city functional area identification system clusters the interest points of the target area by using the spatial clustering mode of automatically determining the number of categories, according to the spatial distribution position of the interest points on the city map and the third-level type of the interest points, to obtain a first clustering result, and divides the functional area into multiple sub-areas, for example, referring to Figure 4 The third-level type of the interest points in the target functional area obtained is 3 types, type A is mostly distributed in the upper left corner of the area, type B is mostly distributed in the lower right corner of the area, and type C is mostly distributed in the lower left corner of the area, so the target functional area is divided into three sub-functional areas, i.e., sub-area 1 (upper left corner), sub-area 2 (lower right corner), and sub-area 3 (lower left).

[0077] S31: determining the topic distribution of each functional sub-region according to the LDA model and all the interest point data of the functional sub-region;

[0078] The LDA model is a probability generation model for a discrete data set, and is a three-layer Bayesian model, in which each element in the set can be regarded as a limited mixture of elements of a set of potential topics.

[0079] The set of the third-level type data of all the interest points in the functional sub-region under the LDA model forms the topic distribution of the functional sub-region. It is easy to understand that there can be multiple topic distributions of each functional sub-region, but the occurrence probabilities of the topic distributions are different. After determining the topic distribution of each functional sub-region, the determined topic distributions can be sorted according to the size of their occurrence probabilities.

[0080] S32: determining the topic whose probability is arranged in the first place in the identified topic distribution as the sub-region semantic feature of the functional sub-region;

[0081] If the preset number is one, the topic whose probability is arranged in the first place, i.e., the topic with the highest occurrence probability, is determined as the sub-region semantic feature of the functional sub-region.

[0082] According to the third-level type data of all the interest points in the functional sub-region under the LDA model, the topic distribution of each functional region is obtained, and the topic whose probability is arranged in the first place in the identified topic distribution is determined as the sub-region semantic feature of the functional sub-region. For example, the topics with the highest probabilities corresponding to three functional sub-regions are shown in Table 1, topic 164 is the topic of the upper functional sub-region, topic 95 and topic 30 are the topics of the lower right functional sub-region, and topic 43 and topic 86 are the topics of the lower left sub-region. The POI types contained in topic 164 and topic 95 are mostly consumer places, and the POI types contained in topic 43 are residences, building numbers and some supporting life services. According to the POI types contained in the topics, it can be inferred that the two functional sub-regions on the upper and right sides of functional region 1 are commercial areas, and the functional sub-region on the lower left is a residential area.

[0083] S41: clustering the sub-region semantic features of all the functional sub-regions to obtain the region space semantic feature of the target functional region;

[0084] That is, the same sub-region semantic features of all the functional sub-regions are merged, and the combination of the at least one sub-region semantic feature finally obtained is the region space semantic feature of the target functional region.

[0085] For example, referring to Table 1, according to the POI types contained in the topic, two functional sub-regions above and right of the target functional region 1 are obtained as commercial areas, and the functional sub-region in the lower left is obtained as a residential area, the sub-region semantic features of the three functional sub-regions are clustered, and the regional spatial semantic feature of the target functional region of the functional region 1 is obtained as a commercial area + a residential area. In this embodiment, the latitude and longitude coordinates of the POI in the target functional region and the third-level type are obtained, the POI is matched to the city map according to the latitude and longitude coordinates of the POI, clustering is performed according to the distribution of the POI on the map to obtain a first clustering result, the target functional region is divided into a plurality of sub-regions according to the first clustering result, the topic distribution of the sub-functional region is obtained according to the third-level type data of all POIs under the LDA model, the main sub-region semantic features of the functional region are determined as the topics whose probabilities are arranged in the front of the preset number of identified topic distributions, and the regional spatial semantic feature of the target functional region is obtained by clustering all sub-region semantic features of all functional sub-regions. The problem that the result of the original LDA model ignores the spatial heterogeneity of the semantic features in the functional region and the influence of a relatively small number of POI types on the regional function, resulting in inaccurate regional identification, is solved, and the accuracy and precision of the functional region identification are improved.

[0086] In order for those skilled in the art to better understand the protection scope of the claims of the present application, the technical solutions described in the claims of the present application are explained and described below through specific implementation examples in specific application scenarios. It can be understood that the following examples are only used to explain the present application, and are not used to limit the protection scope of the claims of the present application.

[0087] Example 1, referring to Figure 4 The LDA topic model and the improved LDA topic model in the present application are used to identify the sub-region semantic features of the functional region 1, respectively, and the topic distributions obtained by the two methods are shown in Table 1:

[0088]

[0089]

[0090]

[0091] As can be seen from Table 1, the topics mined by the LDA topic model are topic 107 and topic 51, the main POI type in topic 107 is company, and the POI type in topic 51 is mainly life service, it can be inferred that the functional type of the functional region 1 is a commercial area.

[0092] Using the method proposed in the present application, the functional region is first divided into three sub-regions, such as Figure 4 (b) As shown in the figure, the three sub-regions correspond to the highest probability topics shown in Table 1, topic 164 is the topic of the upper sub-region, topic 95 and topic 30 are the topics of the lower right sub-region, and topic 43 and topic 86 are the topics of the lower left sub-region. The POI types contained in topic 164 and topic 95 are mostly consumer places, and the POI types contained in topic 43 are residences, building numbers and some supporting life services. According to the POI types contained in the topics, it can be inferred that the upper and right two sub-regions of functional area 1 are commercial areas, and the lower left sub-region is a residential area. Therefore, the functional type of functional area 1 is a commercial area + residential area.

[0093] Comparing the topic results mined by the two different methods, it can be seen that the number of POIs of the commercial type in functional area 1 is relatively large, and the POIs of the commercial type dominate the topics mined by the LDA model, which masks the residential function of the functional area and ignores the influence of a relatively small number of POI classes on the regional function. The main reason for this problem is that the topic model is based on the bag-of-words model and finds topics from the frequency of POI occurrence. According to the spatial distribution of POIs, the present application divides the functional area into sub-regions, so that the residential function in the sub-regions is highlighted.

[0094] Example 2, see Figure 5 The LDA topic model and the improved LDA topic model in the present application are used to identify the semantic features of the sub-regions of functional area 2, respectively, and the topic distributions obtained by the two methods are shown in Table 2:

[0095]

[0096]

[0097]

[0098]

[0099] The results of functional area 2 are shown in Figure 4 and Table 2. The two topics with higher probability obtained by the LDA topic model are Topic 93 and Topic 72, and their POI types are mainly companies and beauty salons, beauty salons and power banks, respectively. Therefore, it can be judged that the functional type of functional area 2 is a commercial area.

[0100] By using the method, the functional area 2 is divided into five sub-areas. The two topics with the highest probability in the upper left area are Topic 164 and Topic 110, and the main POI categories are Chinese restaurant and company, science and culture site and life service site, so it can be judged that the sub-area is a commercial area plus a living area; the two topics with the highest probability in the lower left area are Topic 170 and Topic 133, and the main POI categories are company and building number, Chinese restaurant and company, so it can be judged that the sub-area is a commercial area; the two topics with the highest probability in the upper right area are Topic 146 and Topic 110, and the main POI categories are company and power bank, science and culture site and life service site, so it can be judged that the sub-area is a commercial area plus a living area; the two topics with the highest probability in the middle area are Topic 110 and Topic 35, and the main POI categories are science and culture site and life service site, tourist attraction and science and culture site, so it can be judged that the sub-area is a commercial area plus a living area; the two topics with the highest probability in the lower right area are Topic 146 and Topic 110, and the main POI categories are company and power bank, science and culture site and life service site, so it can be judged that the sub-area is a commercial area plus a living area. It can be judged comprehensively that the functional type of the functional area 2 is a commercial area plus a living area.

[0101] It can be seen that the LDA model is affected by the number of POI categories in the area when extracting the area characteristics. In other words, the results of the LDA model ignore the influence of POI categories with relatively small quantities on the function of the area, and only show the function determined by POI categories with relatively large quantities.

[0102] Therefore, the improved LDA topic model of the present application can more accurately divide the target functional area into multiple sub-areas, and the functions of these sub-areas are more single, so that POI categories with close relationship are always divided together, so that the model can be better trained, and more accurate and detailed results can be obtained.

[0103] Referring to Figure 6 , Figure 6 A structure schematic diagram of a city functional area recognition device based on POI and an improved topic model is provided, and the device specifically comprises:

[0104] A data acquisition module is configured to acquire interest point data of a target functional area, wherein the interest point data comprises spatial position data of each interest point.

[0105] A region division module is configured to divide the target functional area into multiple sub-areas according to the spatial position data.

[0106] A type determining module is configured to determine a sub-region semantic feature of the functional sub-region.

[0107] A type identifying module is configured to obtain a region spatial semantic feature of the target functional region according to the sub-region semantic feature.

[0108] The technical scheme of the embodiment achieves the following effects through the cooperation of the functional modules: the spatial position data of the interest point is obtained by acquiring the interest point data of the target functional region, the target functional region is divided into a plurality of functional sub-regions according to the spatial position data of the interest point, the sub-region semantic feature in each functional sub-region is determined, and the region spatial semantic feature of the target functional region is further obtained, so as to solve the problems of low recognition accuracy and poor accuracy in the existing functional region recognition technology.

[0109] In addition, the embodiment of the present application further provides a computer storage medium, and the computer storage medium stores a city functional region recognition program based on POI and an improved topic model. The city functional region recognition program based on POI and the improved topic model is executed by a processor to implement the steps of the city functional region recognition method based on POI and the improved topic model. Therefore, the description will not be repeated here. In addition, the beneficial effects of the same method will not be described again. For the technical details of the computer readable storage medium embodiments involved in the present application, please refer to the description of the method embodiments of the present application. For example, the program instructions can be deployed to be executed on one computing device, or on multiple computing devices located in one place, or on multiple computing devices distributed in multiple places and interconnected through a communication network.

[0110] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by a computer program instructing related hardware. The above-mentioned program can be stored in a computer readable storage medium, and when the program is executed, it can include the processes of the above-mentioned method embodiments. The storage medium can be a magnetic disc, an optical disc, a read-only memory (ROM) or a random access memory (RAM), etc.

[0111] In addition, it should be noted that the apparatus embodiments described above are merely illustrative, and units described as separate components can or can not be physically separate, and components shown as units can or can not be physical units, i.e., can be located in one place, or can be distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the embodiments. In addition, the connection between the modules in the apparatus embodiments provided in the present application indicates that there is a communication connection between them, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement it without creative labor.

[0112] Through the description of the above embodiments, those skilled in the art can clearly understand that the present application can be realized by means of software and necessary general hardware, and of course can also be realized by special hardware including special integrated circuits, special CPUs, special memories, special components, etc. Generally, functions completed by computer programs can be easily realized by corresponding hardware, and specific hardware structures for realizing the same function can also be various, such as analog circuits, digital circuits or special circuits. However, for the present application, software program implementation is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of software products, which are stored in readable storage media, such as computer floppy disks, U disks, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks or optical disks, etc., including a number of instructions to make a computer device (which can be a personal computer, server, or network device, etc.) execute the methods of various embodiments of the present application.

[0113] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application, and any equivalent structure or equivalent process transformation using the content of the present application specification and drawings, or direct or indirect application in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for identifying urban functional zones based on Points of Interest (POI) and an improved topic model, characterized in that, The method includes: Acquire point-of-interest (POI) data for the target functional area, wherein the POI data includes spatial location data and third-level type data for each POI; Based on the spatial location data and the third-level type data, the target functional area is divided into multiple functional sub-areas; Determine the sub-region semantic features of each of the aforementioned functional sub-regions; Based on the semantic features of the sub-regions, the regional spatial semantic features of the target functional region are obtained; The step of dividing the target functional area into multiple functional sub-regions based on the spatial location data and the third-level type data includes: Based on the spatial location data and the third-level type data, the points of interest are clustered to obtain the first clustering result; Based on the first clustering result, the target functional region is divided into multiple sub-regions.

2. The urban functional area identification method based on POI and improved topic model according to claim 1, characterized in that, The acquisition of interest point data for the target functional area, wherein the interest point data includes spatial location data of each interest point, including: Obtain urban point of interest (POI) data of the city where the target functional area is located, wherein the POI data includes spatial location data of the POIs; Based on the spatial location data of the points of interest, the city point of interest data is matched onto the city map to obtain point of interest data for the target functional area.

3. The urban functional area identification method based on POI and improved topic model according to claim 1, characterized in that, Determining the sub-region semantic features of each of the functional sub-regions includes: Based on the LDA model and all interest point data of the functional sub-regions, determine the topic distribution of each functional sub-region; The topics with the highest probability distribution among the identified topics are determined as the semantic features of the sub-regions of the functional sub-regions.

4. The urban functional area identification method based on POI and improved topic model according to any one of claims 1 to 3, characterized in that, The step of obtaining the regional spatial semantic features of the target functional region based on the semantic features of the sub-region includes: Cluster the semantic features of all the functional sub-regions to obtain the feature type of the target functional region.

5. The urban functional area identification method based on POI and improved topic model according to any one of claims 1-4, characterized in that, include: If the semantic features of each of the functional sub-regions belong to the same type, then the semantic features of the sub-regions are used as the regional spatial semantic features of the target functional region.

6. The urban functional area identification method based on POI and improved topic model according to any one of claims 1-4, characterized in that, include: If the semantic features of the sub-regions of each of the functional sub-regions do not belong to the same type, then the semantic features of the sub-regions of all the functional sub-regions are clustered to obtain the second clustering result; Based on the second clustering result, the regional spatial semantic features of the target functional region are determined.

7. A device for identifying urban functional areas based on Points of Interest (POIs) and an improved topic model, characterized in that, The device includes: The data acquisition module is used to acquire point-of-interest (POI) data of the target functional area, wherein the POI data includes spatial location data of each POI and third-level type data; The region division module is used to divide the target functional area into multiple sub-regions based on the spatial location data and the third-level type data; A type determination module is used to determine the sub-region semantic features of the functional sub-region; The type recognition module is used to obtain the regional spatial semantic features of the target functional region based on the semantic features of the sub-region; The region division module is further configured to cluster the points of interest based on the spatial location data and the third-level type data to obtain a first clustering result; and to divide the target functional region into multiple sub-regions based on the first clustering result.

8. A device for identifying urban functional areas based on POIs and an improved topic model, characterized in that, include: A processor, a memory, and a POI-based and improved topic model-based urban functional area identification program stored in the memory, wherein the POI-based and improved topic model-based urban functional area identification program is executed by the processor to implement the steps of the POI-based and improved topic model-based urban functional area identification method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program for identifying urban functional zones based on POI and an improved topic model, which, when executed by a processor, implements the method for identifying urban functional zones based on POI and an improved topic model as described in any one of claims 1 to 6.

Images