Method for determining spatiotemporal co-occurrence zones, non-public places and user social behavior relationships

Abstract

The application discloses a method for determining spatio-temporal co-occurrence zones, non-public places and social behavior relationships of users, and belongs to the technical field of data processing. The method first divides the spatio-temporal co-occurrence zones by using location data interest points, then calculates the location entropy of each interest point in sequence to complete the identification of the non-public places, and finally screens the spatio-temporal co-occurrence zones according to the non-public places to determine the social behavior relationships of the users existing in the spatio-temporal co-occurrence zones. The scheme solves the problems that the spatio-temporal co-occurrence zone division is inaccurate, the public place interest points pollute the overall data, the social relationship inference accuracy is not high, and the network space social behavior relationship data completion quality is low.

Description

Technical Field

[0001] This invention belongs to the field of data processing technology, and in particular relates to a method for determining spatiotemporal co-occurrence zones, non-public places, and the social behavior relationships of users. Background Technology

[0002] Cyberspace mapping involves detecting, collecting, identifying, analyzing, and processing the attributes and relationships of cyberspace facilities, services, and resources. Based on geographic information and logical relationships, it creates maps that comprehensively describe and display cyberspace assets, attributes, status, relationships, and trends, providing data and technical support for various information security applications.

[0003] Cyberspace consists of several distinct but interconnected layers (physical network layer, logical network layer, and social role layer), each capturing key characteristics and behaviors within the domain. The social role layer describes the composition and characteristics of social networks, including the attributes of social actors and their relationships. Social actors are user roles that manage, operate, and use various network devices and application services, while the edges between social actors represent the social relationships between users.

[0004] Social role layer data is primarily obtained through social media, social engineering, and open-source intelligence. However, current social role layer data mainly consists of discrete nodes, lacking social actor relationships (user social relationships). With the development of LBS technology, location information has been added to some social actor attributes. Mining and analyzing location data has become a feasible approach to supplementing social role layer relationship data in cyberspace mapping.

[0005] Currently, social relationship inference based on location data mainly uses two methods: (1) establishing spatio-temporal co-occurrence zones for user visits and inferring relationships based on the frequency and other relevant characteristics of these zones, combined with machine learning methods (such as random forests); (2) judging social relationships based on the similarity of user trajectories. Since the second method requires a high degree of data continuity, the spatio-temporal co-occurrence zone method is more commonly used.

[0006] The so-called "spatiotemporal co-occurrence" refers to two users, u1 and u2, existing at the same time. They all appeared in the same area Spatiotemporal co-occurrence region extraction uses two representations of dataset D as input (first form: coordinate dataset). The second form: Region ID dataset D = {(u i ,t i ,l i )|i=1,2,…,W}), by partitioning and The process of extracting relevant users.

[0007] However, current inference methods based on spatiotemporal co-occurrence zones have insufficient accuracy and efficiency, mainly due to two problems: (1) Classic spatiotemporal co-occurrence zones do not consider the circular distribution of user visit locations and are divided using fixed grids, leading to users belonging to the same location being mistakenly assigned to other areas; (2) Spatiotemporal co-occurrence zones contain public places (such as shopping malls and stations) that can be visited by all users, and using this type of data for social relationship inference may lead to a decrease in accuracy. These two problems ultimately result in low quality of social role layer data completion. Summary of the Invention

[0008] To address the aforementioned technical problems, this invention proposes a scheme for determining the spatiotemporal co-occurrence zone, non-public places, and the relationship between user social behavior.

[0009] The first aspect of this invention discloses a method for determining a spatiotemporal co-occurrence region. The method includes:

[0010] Step S1: Use the user's location dataset to determine the area range of each location data point contained in the location dataset, and extract K points of interest from the area range;

[0011] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0012] Step S2: Based on the center coordinates of the K points of interest, determine L candidate location data from the M location data, and associate the L candidate location data with the K points of interest;

[0013] Step S3: For each of the K points of interest, obtain at least two candidate location data within a preset time period from the candidate location data associated with it, and further determine at least two users to which the at least two candidate location data belong, and determine the area containing the current point of interest and the at least two candidate location data as the spatiotemporal co-occurrence zone between the at least two users.

[0014] According to the method of the first aspect of the present invention, in step S1, determining the region range containing each location data in the location dataset includes: marking each location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region range, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0015] According to the method of the first aspect of the present invention, in step S2:

[0016] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0017] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0018] According to the method of the first aspect of the present invention, in step S3:

[0019] For each of the K points of interest, acquire the candidate location data associated with it, and determine whether the number of candidate location data existing within the preset time period is greater than or equal to two. If so, extract the at least two candidate location data.

[0020] For the extracted at least two candidate location data, determine the users to which the at least two candidate location data belong, and determine whether the number of users to which they belong is greater than or equal to two. If so, determine that there is a spatiotemporal co-occurrence zone between the at least two users.

[0021] A second aspect of the present invention discloses a method for identifying non-public places, the method comprising:

[0022] Step S1: Use the user's location dataset to determine the area range of each location data point contained in the location dataset, and extract K points of interest from the area range;

[0023] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0024] Step S2: Based on the center coordinates of the K points of interest, determine L candidate location data from the M location data, and associate the L candidate location data with the K points of interest;

[0025] Step S3: For each of the K points of interest, calculate the location entropy of the current point of interest based on the candidate location data associated with it. When the location entropy is less than the entropy threshold, determine that the current point of interest is the non-public place.

[0026] The position entropy is calculated using the following formula:

[0027]

[0028]

[0029] in, For user u j Visit the current point of interest z i Number of times, For all users in the user set U to visit the current point of interest z i Total number of times For the user u j Visit the current point of interest z i The percentage of users who visited that location.

[0030] According to the method of the second aspect of the present invention, in step S1, determining the region range containing each location data in the location dataset includes: marking each location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region range, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0031] According to the method of the second aspect of the present invention, in step S2:

[0032] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0033] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0034] A third aspect of this invention discloses a method for determining user social behavior relationships based on spatiotemporal co-occurrence regions, the method comprising:

[0035] Step S1: Use the user's location dataset to determine the area range of each location data point contained in the location dataset, and extract K points of interest from the area range;

[0036] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0037] Step S2: Based on the center coordinates of the K points of interest, determine L candidate location data from the M location data, and associate the L candidate location data with the K points of interest;

[0038] Step S3: For each of the K points of interest, obtain at least two candidate location data within a preset time period from the candidate location data associated with it, and further determine at least two users to which the at least two candidate location data belong, and determine the area containing the current point of interest and the at least two candidate location data as the spatiotemporal co-occurrence area between the at least two users.

[0039] Step S4: For each of the K points of interest, calculate the location entropy of the current point of interest based on the candidate location data associated with it. When the location entropy is less than the entropy threshold, determine that the current point of interest is the non-public place.

[0040] The position entropy is calculated using the following formula:

[0041]

[0042]

[0043] in, For user u j Visit the current point of interest z i Number of times, For all users in the user set U to visit the current point of interest z i Total number of times For the user u j Visit the current point of interest z i The percentage of users who visited that location.

[0044] Step S5: Extract the points of interest and associated users contained in the spatiotemporal co-occurrence zone. When the contained points of interest are non-public places, determine the social behavior relationships between the associated users based on the non-public places.

[0045] According to the method of the third aspect of the present invention, in step S1, determining the region range containing each location data in the location dataset includes: marking each location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region range, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0046] According to the method of the third aspect of the present invention, in step S2:

[0047] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0048] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0049] According to the method of the third aspect of the present invention, in step S3:

[0050] For each of the K points of interest, acquire the candidate location data associated with it, and determine whether the number of candidate location data existing within the preset time period is greater than or equal to two. If so, extract the at least two candidate location data.

[0051] For the extracted at least two candidate location data, determine the users to which the at least two candidate location data belong, and determine whether the number of users to which they belong is greater than or equal to two. If so, determine that there is a spatiotemporal co-occurrence zone between the at least two users.

[0052] A fourth aspect of the present invention discloses a system for determining a spatiotemporal co-occurrence region, the system comprising,

[0053] The first processing unit is configured to: use the user's location dataset to determine the regional range of each location data contained in the location dataset, and extract K points of interest from the regional range;

[0054] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0055] The second processing unit is configured to: determine L candidate location data from the M location data based on the center coordinates of the K points of interest, and associate the L candidate location data with the K points of interest;

[0056] The third processing unit is configured to: for each of the K points of interest, acquire at least two candidate location data within a preset time period from the candidate location data associated with it, and further determine at least two users to which the at least two candidate location data belong, and determine the area containing the current point of interest and the at least two candidate location data as the spatiotemporal co-occurrence zone between the at least two users.

[0057] According to the system of the fourth aspect of the present invention, the first processing unit is specifically configured to: determine the region range containing each piece of location data in the location dataset by: marking each piece of location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each piece of location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region range, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0058] According to the system of the fourth aspect of the present invention, the second processing unit is specifically configured as follows:

[0059] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0060] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0061] According to the system of the fourth aspect of the present invention, the third processing unit is specifically configured as follows:

[0062] For each of the K points of interest, acquire the candidate location data associated with it, and determine whether the number of candidate location data existing within the preset time period is greater than or equal to two. If so, extract the at least two candidate location data.

[0063] For the extracted at least two candidate location data, determine the users to which the at least two candidate location data belong, and determine whether the number of users to which they belong is greater than or equal to two. If so, determine that there is a spatiotemporal co-occurrence zone between the at least two users.

[0064] A fifth aspect of the present invention discloses a system for identifying non-public places, the system comprising:

[0065] The first processing unit is configured to: use the user's location dataset to determine the regional range of each location data contained in the location dataset, and extract K points of interest from the regional range;

[0066] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0067] The second processing unit is configured to: determine L candidate location data from the M location data based on the center coordinates of the K points of interest, and associate the L candidate location data with the K points of interest;

[0068] The third processing unit is configured to: for each of the K points of interest, calculate the location entropy of the current point of interest based on the candidate location data associated with it; and when the location entropy is less than the entropy threshold, determine that the current point of interest is the non-public place.

[0069] The position entropy is calculated using the following formula:

[0070]

[0071]

[0072] in, For user u j Visit the current point of interest z i Number of times, For all users in the user set U to visit the current point of interest z i Total number of times For the user u j Visit the current point of interest z i The percentage of users who visited that location.

[0073] According to the system of the fifth aspect of the present invention, the first processing unit is specifically configured to: determine the region range containing each piece of location data in the location dataset by: marking each piece of location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each piece of location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region range, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0074] According to the system of the fifth aspect of the present invention, the second processing unit is specifically configured as follows:

[0075] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0076] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0077] The sixth aspect of this invention discloses a system for determining user social behavior relationships based on spatiotemporal co-occurrence zones, the system comprising:

[0078] The first processing unit is configured to: use the user's location dataset to determine the regional range of each location data contained in the location dataset, and extract K points of interest from the regional range;

[0079] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0080] The second processing unit is configured to: determine L candidate location data from the M location data based on the center coordinates of the K points of interest, and associate the L candidate location data with the K points of interest;

[0081] The third processing unit is configured to: for each of the K points of interest, acquire at least two candidate location data that are within a preset time period from the candidate location data associated with it, and further determine at least two users to which the at least two candidate location data belong, and determine the area containing the current point of interest and the at least two candidate location data as the spatiotemporal co-occurrence zone between the at least two users;

[0082] The fourth processing unit is configured to: calculate the location entropy of each of the K points of interest based on the candidate location data associated with it; and determine that the current point of interest is the non-public place when the location entropy is less than the entropy threshold.

[0083] The position entropy is calculated using the following formula:

[0084]

[0085]

[0086] in, For user u j Visit the current point of interest z i Number of times, For all users in the user set U to visit the current point of interest z i Total number of times For the user u j Visit the current point of interest z i The percentage of users who visited that location.

[0087] The fifth processing unit is configured to: extract points of interest and associated users contained in the spatiotemporal co-occurrence zone; and when the included points of interest are non-public places, determine the social behavioral relationships between the associated users based on the non-public places.

[0088] According to the system of the sixth aspect of the present invention, the first processing unit is specifically configured to: determine the region range containing each piece of location data in the location dataset by: marking each piece of location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each piece of location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region range, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0089] According to the system of the sixth aspect of the present invention, the second processing unit is specifically configured as follows:

[0090] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0091] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0092] According to the system of the sixth aspect of the present invention, the third processing unit is specifically configured as follows:

[0093] For each of the K points of interest, acquire the candidate location data associated with it, and determine whether the number of candidate location data existing within the preset time period is greater than or equal to two. If so, extract the at least two candidate location data.

[0094] For the extracted at least two candidate location data, determine the users to which the at least two candidate location data belong, and determine whether the number of users to which they belong is greater than or equal to two. If so, determine that there is a spatiotemporal co-occurrence zone between the at least two users.

[0095] A seventh aspect of the present invention discloses a computer-readable storage medium. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of any of the methods of the first, second, and third aspects of this disclosure.

[0096] An eighth aspect of the present invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of any one of the methods of the first, second, and third aspects of this disclosure.

[0097] In summary, the technical solutions provided by this invention offer a method for dividing spatiotemporal co-occurrence zones to address the problem that the existing methods for dividing spatiotemporal co-occurrence zones using grid regions are not accurate enough; a method for identifying public places using location entropy; and a method for completing social behavior relationship data based on location data to achieve high-precision completion of social behavior relationship data. Attached Figure Description

[0098] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0099] Figure 1 A flowchart for determining the spatiotemporal co-occurrence region according to Embodiment 1 of the present invention;

[0100] Figure 2a and 2b A comparison chart of spatiotemporal co-occurrence region extraction methods;

[0101] Figure 3 A flowchart for determining non-public places according to Embodiment 2 of the present invention;

[0102] Figure 4 A flowchart for determining social behavioral relationships according to Embodiment 3 of the present invention;

[0103] Figure 5 This is a structural diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0104] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0105] The first aspect of this invention discloses a method for determining a spatiotemporal co-occurrence region. The method includes:

[0106] Step S1: Use the user's location dataset to determine the area range of each location data point contained in the location dataset, and extract K points of interest from the area range;

[0107] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0108] Step S2: Based on the center coordinates of the K points of interest, determine L candidate location data from the M location data, and associate the L candidate location data with the K points of interest;

[0109] Step S3: For each of the K points of interest, obtain at least two candidate location data within a preset time period from the candidate location data associated with it, and further determine at least two users to which the at least two candidate location data belong, and determine the area containing the current point of interest and the at least two candidate location data as the spatiotemporal co-occurrence zone between the at least two users.

[0110] Note that after determining the existence of a spatiotemporal co-occurrence zone, the scope of the spatiotemporal co-occurrence zone can be determined not only by "determining the area containing the current point of interest and the at least two candidate location data as the spatiotemporal co-occurrence zone between the at least two users" in step S3, but also by other delineation methods in this field. For example, a circular area can be drawn with the center coordinates of the current point of interest as the center and a preset value (500 meters, 1000 meters, etc.) as the radius as the spatiotemporal co-occurrence zone.

[0111] In some embodiments, in step S1, determining the region containing each location data in the location dataset includes: marking each location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0112] In some embodiments, in step S2:

[0113] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0114] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0115] In some embodiments, in step S3:

[0116] For each of the K points of interest, acquire the candidate location data associated with it, and determine whether the number of candidate location data existing within the preset time period is greater than or equal to two. If so, extract the at least two candidate location data.

[0117] For the extracted at least two candidate location data, determine the users to which the at least two candidate location data belong, and determine whether the number of users to which they belong is greater than or equal to two. If so, determine that there is a spatiotemporal co-occurrence zone between the at least two users.

[0118] Specific Implementation Example 1 (as shown) Figure 1 As shown):

[0119] S01: Input coordinate location dataset or region ID location dataset;

[0120] S02: Determine the dataset type. If the dataset type is coordinate data, execute S03. If the dataset type is region ID data, execute S07.

[0121] S03: Based on the range of all user location sets, query the geodatabase for POIs (Point of Interest, semantic locations) within that location range, and extract the center coordinates corresponding to the POIs in sequence;

[0122] S04: Set the spatiotemporal co-occurrence zone discrimination threshold λ, and calculate the distance between the center coordinates of each POI and the user coordinates in turn;

[0123] S05: Determine whether the distance between the center coordinates of the POI and the user coordinates is less than the threshold. If it is less than the threshold, execute S06; otherwise, execute S03 to extract the center coordinates of the POI in sequence.

[0124] S06: Add the data points to the candidate set according to the POI identifier, and then execute S09;

[0125] S07: If the dataset type is region ID data, based on the user's region ID set, iterate through and query the set of points of interest (POIs) in the geographic database that correspond to the region ID set, and extract the region IDs corresponding to the POIs in turn.

[0126] S08: Determine whether the POI's corresponding region ID is consistent with the user's region ID. If they are consistent, proceed to S09; otherwise, proceed to S07 to sequentially select and extract the POI's corresponding region ID.

[0127] S09: Add the data points to the candidate location set according to the POI identifier;

[0128] S10: Select the candidate set of location data arranged by POI identifier in sequence;

[0129] S11: Delineation For a time interval of τ, determine whether each data point in the candidate set has more than two location data points within the time interval τ. If so, execute S12; otherwise, execute S10.

[0130] S12: Determine whether there are more than 2 users in the candidate set that meet the conditions in S11. If so, the candidate set is a spatiotemporal co-occurrence zone; otherwise, execute S11.

[0131] S13: Output the spatiotemporal co-occurrence zone identifier and user ID.

[0132] like Figure 2a and 2b As shown, the original spatiotemporal co-occurrence region identification method uses regions Let the grid be λ×λ squares, and assume users u1 and u2 are in... All appeared in the region but The new spatiotemporal co-occurrence region extraction method divides and identifies regions based on a circular area with radius λ. Comparison reveals that the new spatiotemporal co-occurrence region identification method is more accurate.

[0133] A second aspect of the present invention discloses a method for identifying non-public places, the method comprising:

[0134] Step S1: Use the user's location dataset to determine the area range of each location data point contained in the location dataset, and extract K points of interest from the area range;

[0135] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0136] Step S2: Based on the center coordinates of the K points of interest, determine L candidate location data from the M location data, and associate the L candidate location data with the K points of interest;

[0137] Step S3: For each of the K points of interest, calculate the location entropy of the current point of interest based on the candidate location data associated with it. When the location entropy is less than the entropy threshold, determine that the current point of interest is the non-public place.

[0138] The position entropy is calculated using the following formula:

[0139]

[0140]

[0141] in, For user u j Visit the current point of interest z i Number of times, For all users in the user set U to visit the current point of interest z i Total number of times For the user u j Visit the current point of interest z i The percentage of users who visited that location.

[0142] According to the method of the second aspect of the present invention, in step S1, determining the region range containing each location data in the location dataset includes: marking each location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region range, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0143] According to the method of the second aspect of the present invention, in step S2:

[0144] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0145] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0146] Specific Implementation Example 2 (as shown) Figure 3 As shown):

[0147] Because public places have a large number of users and are highly random, the accuracy of relationship inference using this spatiotemporal data is low. To avoid contaminating the overall dataset, this part of the data needs to be removed. Therefore, it is necessary to identify data from non-public places.

[0148] To address this issue, a location entropy index is used to separate public and non-public locations. A higher location entropy value indicates greater information gain, lower predictability, and more active user activity, suggesting a higher probability that the location is a public place. Conversely, a lower location entropy value indicates less information gain, better predictability, and lower user activity, suggesting a higher probability that the location is a non-public place.

[0149] The specific methods and procedures include:

[0150] S01: Input coordinate location dataset or region ID location dataset;

[0151] S02: Determine the dataset type. If the dataset type is coordinate data, execute S03. If the dataset type is region ID data, execute S07.

[0152] S03: Based on the range of all user location sets, query the geodatabase for POIs (Point of Interest, semantic locations) within that location range, and extract the center coordinates corresponding to the POIs in sequence;

[0153] S04: Set the spatiotemporal co-occurrence zone discrimination threshold λ, and calculate the distance between the center coordinates of each POI and the user coordinates in turn;

[0154] S05: Determine whether the distance between the center coordinates of the POI and the user coordinates is less than the threshold. If it is less than the threshold, execute S06; otherwise, execute S03 to extract the center coordinates of the POI in sequence.

[0155] S06: Add the data points to the candidate set according to the POI identifier, and then execute S09;

[0156] S07: If the dataset type is region ID data, based on the user's region ID set, iterate through and query the set of points of interest (POIs) in the geographic database that correspond to the region ID set, and extract the region IDs corresponding to the POIs in turn.

[0157] S08: Determine whether the POI's corresponding region ID is consistent with the user's region ID. If they are consistent, proceed to S09; otherwise, proceed to S07 to sequentially select and extract the POI's corresponding region ID.

[0158] S09: Add the data points to the candidate set according to the POI identifier;

[0159] S10: Select the candidate set of location data arranged by POI identifier in sequence;

[0160] S11: Set the location entropy threshold α, and use the candidate location set corresponding to each POI as input to calculate the location entropy of the area corresponding to the POI according to the following formula (used to evaluate the predictability of the user in the area).

[0161]

[0162]

[0163] in, For user u j Visited area z i Number of times, For the visited location z i Total quantity For user u j Visited area z i The percentage of users who visited that location.

[0164] S12: Judgment Is it less than α? If the value is less than α, then execute S13; otherwise, execute S10.

[0165] S13: Identify as a non-public place, record the POI identifier and output it.

[0166] A third aspect of this invention discloses a method for determining user social behavior relationships based on spatiotemporal co-occurrence regions, the method comprising:

[0167] Step S1: Use the user's location dataset to determine the area range of each location data point contained in the location dataset, and extract K points of interest from the area range;

[0168] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0169] Step S2: Based on the center coordinates of the K points of interest, determine L candidate location data from the M location data, and associate the L candidate location data with the K points of interest;

[0170] Step S3: For each of the K points of interest, obtain at least two candidate location data within a preset time period from the candidate location data associated with it, and further determine at least two users to which the at least two candidate location data belong, and determine the area containing the current point of interest and the at least two candidate location data as the spatiotemporal co-occurrence area between the at least two users.

[0171] Step S4: For each of the K points of interest, calculate the location entropy of the current point of interest based on the candidate location data associated with it. When the location entropy is less than the entropy threshold, determine that the current point of interest is the non-public place.

[0172] The position entropy is calculated using the following formula:

[0173]

[0174]

[0175] in, For user u j Visit the current point of interest z i Number of times, For all users in the user set U to visit the current point of interest z i Total number of times For the user u j Visit the current point of interest z i The percentage of users who visited that location.

[0176] Step S5: Extract the points of interest and associated users contained in the spatiotemporal co-occurrence zone. When the contained points of interest are non-public places, determine the social behavior relationships between the associated users based on the non-public places.

[0177] According to the method of the third aspect of the present invention, in step S1, determining the region range containing each location data in the location dataset includes: marking each location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region range, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0178] According to the method of the third aspect of the present invention, in step S2:

[0179] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0180] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0181] According to the method of the third aspect of the present invention, in step S3:

[0182] For each of the K points of interest, acquire the candidate location data associated with it, and determine whether the number of candidate location data existing within the preset time period is greater than or equal to two. If so, extract the at least two candidate location data.

[0183] For the extracted at least two candidate location data, determine the users to which the at least two candidate location data belong, and determine whether the number of users to which they belong is greater than or equal to two. If so, determine that there is a spatiotemporal co-occurrence zone between the at least two users.

[0184] Specific embodiment 3 (as shown) Figure 4 As shown):

[0185] S01: Input coordinate location dataset or region ID location dataset;

[0186] S02: Extract the spatiotemporal co-occurrence region and user ID according to the steps of the first aspect of the present invention;

[0187] S03: Extract non-public place POI identifiers according to the steps of the second aspect of the present invention, and generate a non-public place POI identifier library;

[0188] S04: Select a spatiotemporal co-occurrence region in sequence;

[0189] S05: Determine whether the POI to which the spatiotemporal co-occurrence zone belongs is included in the POI library generated in step S03. If it is included, proceed to S06; otherwise, proceed to S04.

[0190] S06: Determine that there are social relationships among users in the spatiotemporal co-occurrence zone, and complete the social actor relationship data.

[0191] In some embodiments, point of interest A and its associated users u1 and u2 are extracted from the spatiotemporal co-occurrence zone. If point of interest A is a non-public place, it can be determined that u1 and user u2 have a social relationship. Point of interest B and its associated users u3 and u4 are extracted from the spatiotemporal co-occurrence zone. If point of interest B is a non-public place, it can be determined that u3 and user u4 have a social relationship.

[0192] Specifically, the extracted point of interest A is used as a social attribute of a non-public place to establish the social behavioral relationship between user u1 and user u2; for example, if A is an office building, it can be inferred that user u1 and user u2 belong to the same office area and are colleagues or have a work-related relationship.

[0193] Specifically, the extracted point of interest B is used as a social attribute of a non-public place to establish the social behavioral relationship between user u3 and user u4. For example, if B is a dormitory building on a university campus, it can be inferred that user u3 and user u4 live in the same dormitory area and are classmates or a student and administrator in the same dormitory building.

[0194] A fourth aspect of the present invention discloses a system for determining a spatiotemporal co-occurrence region, the system comprising,

[0195] The first processing unit is configured to: use the user's location dataset to determine the regional range of each location data contained in the location dataset, and extract K points of interest from the regional range;

[0196] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0197] The second processing unit is configured to: determine L candidate location data from the M location data based on the center coordinates of the K points of interest, and associate the L candidate location data with the K points of interest;

[0198] The third processing unit is configured to: for each of the K points of interest, acquire at least two candidate location data within a preset time period from the candidate location data associated with it, and further determine at least two users to which the at least two candidate location data belong, and determine the area containing the current point of interest and the at least two candidate location data as the spatiotemporal co-occurrence zone between the at least two users.

[0199] According to the system of the fourth aspect of the present invention, the first processing unit is specifically configured to: determine the region range containing each piece of location data in the location dataset by: marking each piece of location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each piece of location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region range, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0200] According to the system of the fourth aspect of the present invention, the second processing unit is specifically configured as follows:

[0201] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0202] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0203] According to the system of the fourth aspect of the present invention, the third processing unit is specifically configured as follows:

[0204] For each of the K points of interest, acquire the candidate location data associated with it, and determine whether the number of candidate location data existing within the preset time period is greater than or equal to two. If so, extract the at least two candidate location data.

[0205] For the extracted at least two candidate location data, determine the users to which the at least two candidate location data belong, and determine whether the number of users to which they belong is greater than or equal to two. If so, determine that there is a spatiotemporal co-occurrence zone between the at least two users.

[0206] A fifth aspect of the present invention discloses a system for identifying non-public places, the system comprising:

[0207] The first processing unit is configured to: use the user's location dataset to determine the regional range of each location data contained in the location dataset, and extract K points of interest from the regional range;

[0208] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0209] The second processing unit is configured to: determine L candidate location data from the M location data based on the center coordinates of the K points of interest, and associate the L candidate location data with the K points of interest;

[0210] The third processing unit is configured to: for each of the K points of interest, calculate the location entropy of the current point of interest based on the candidate location data associated with it; and when the location entropy is less than the entropy threshold, determine that the current point of interest is the non-public place.

[0211] The position entropy is calculated using the following formula:

[0212]

[0213]

[0214] in, For user u j Visit the current point of interest z i Number of times, For all users in the user set U to visit the current point of interest z i Total number of times For the user u j Visit the current point of interest z i The percentage of users who visited that location.

[0215] According to the system of the fifth aspect of the present invention, the first processing unit is specifically configured to: determine the region range containing each piece of location data in the location dataset by: marking each piece of location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each piece of location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region range, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0216] According to the system of the fifth aspect of the present invention, the second processing unit is specifically configured as follows:

[0217] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0218] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0219] The sixth aspect of this invention discloses a system for determining user social behavior relationships based on spatiotemporal co-occurrence zones, the system comprising:

[0220] The first processing unit is configured to: use the user's location dataset to determine the regional range of each location data contained in the location dataset, and extract K points of interest from the regional range;

[0221] The location dataset contains M location data points for N users, where the location data is either the coordinate data of locations visited by the user or the identification data of areas visited by the user; the points of interest are semantically represented hotspots on the map, and each point of interest has its center coordinates.

[0222] The second processing unit is configured to: determine L candidate location data from the M location data based on the center coordinates of the K points of interest, and associate the L candidate location data with the K points of interest;

[0223] The third processing unit is configured to: for each of the K points of interest, acquire at least two candidate location data that are within a preset time period from the candidate location data associated with it, and further determine at least two users to which the at least two candidate location data belong, and determine the area containing the current point of interest and the at least two candidate location data as the spatiotemporal co-occurrence zone between the at least two users;

[0224] The fourth processing unit is configured to: calculate the location entropy of each of the K points of interest based on the candidate location data associated with it; and determine that the current point of interest is the non-public place when the location entropy is less than the entropy threshold.

[0225] The position entropy is calculated using the following formula:

[0226]

[0227]

[0228] in, For user u j Visit the current point of interest z i Number of times, For all users in the user set U to visit the current point of interest z i Total number of times For the user u j Visit the current point of interest z i The percentage of users who visited that location.

[0229] The fifth processing unit is configured to: extract points of interest and associated users contained in the spatiotemporal co-occurrence zone; and when the included points of interest are non-public places, determine the social behavioral relationships between the associated users based on the non-public places.

[0230] According to the system of the sixth aspect of the present invention, the first processing unit is specifically configured to: determine the region range containing each piece of location data in the location dataset by: marking each piece of location data in a two-dimensional coordinate system, determining the maximum and minimum points of the horizontal axis and the maximum and minimum points of the vertical axis of each piece of location data in the two-dimensional coordinate system, and using the above four points to form a rectangular region as the region range, wherein the sides of the rectangular region are parallel or perpendicular to the coordinate axes of the two-dimensional coordinate system.

[0231] According to the system of the sixth aspect of the present invention, the second processing unit is specifically configured as follows:

[0232] When the location data is the coordinate data of the location visited by the user, for each of the K points of interest, the distance between each coordinate data and the center coordinate of the current point of interest is calculated. When the distance is less than the distance threshold, the coordinate data is determined to be the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0233] When the location data is the identifier data of the area visited by the user, for each of the K points of interest, determine all area identifiers containing its center coordinates, and determine whether the identifier data of the area visited by the user is among all the area identifiers. If so, use the identifier data as the candidate location data of the current point of interest, so as to associate the candidate location data with the current point of interest.

[0234] According to the system of the sixth aspect of the present invention, the third processing unit is specifically configured as follows:

[0235] For each of the K points of interest, acquire the candidate location data associated with it, and determine whether the number of candidate location data existing within the preset time period is greater than or equal to two. If so, extract the at least two candidate location data.

[0236] For the extracted at least two candidate location data, determine the users to which the at least two candidate location data belong, and determine whether the number of users to which they belong is greater than or equal to two. If so, determine that there is a spatiotemporal co-occurrence zone between the at least two users.

[0237] A seventh aspect of the present invention discloses a computer-readable storage medium. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of any of the methods of the first, second, and third aspects of this disclosure.

[0238] An eighth aspect of the present invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of any one of the methods of the first, second, and third aspects of this disclosure.

[0239] Figure 5 This is a structural diagram of an electronic device according to an embodiment of the present invention, such as... Figure 5As shown, the electronic device includes a processor, memory, communication interface, display screen, and input device connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, Near Field Communication (NFC), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input device can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the device's casing, or an external keyboard, touchpad, or mouse.

[0240] Those skilled in the art will understand that Figure 5 The structure shown is merely a structural diagram of the part related to the technical solution of this disclosure and does not constitute a limitation on the electronic device to which the solution of this application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0241] In summary, the technical solutions provided by this invention offer a method for dividing spatiotemporal co-occurrence zones to address the problem that the existing methods for dividing spatiotemporal co-occurrence zones using grid regions are not accurate enough; a method for identifying public places using location entropy; and a method for completing social behavior relationship data based on location data to achieve high-precision completion of social behavior relationship data.

[0242] According to the above-mentioned scheme of the present invention, the following technical effects are achieved: (1) The new spatiotemporal co-occurrence zone division method is used to achieve accurate division of the spatiotemporal co-occurrence zone, laying a data foundation for subsequent social relationship inference and discrimination; (2) The non-public place identification method based on location entropy is used to identify non-public places such as workplaces and residences, thereby achieving separation from the public place location data that restricts the accuracy of social relationship inference; (3) A social actor relationship data completion method based on location data is proposed to achieve high-precision inference of social relationships, thereby ultimately achieving social actor relationship data completion of the social role layer in cyberspace.

[0243] Please note that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments have been described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification. The above embodiments only illustrate several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be pointed out that for those skilled in the art, several modifications and improvements can be made without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method of determining a non-public place, characterized in that, The method comprises: Step S1, determining a region range of each piece of location data contained in a location data set of a user by using the location data set, and extracting K interest points in the region range; Wherein, the location data set contains M pieces of location data of N users, the location data is coordinate data of a location visited by the user or identification data of a region visited by the user; the interest point is a semantic hot spot position on a map, and each interest point has a center coordinate; Step S2, determining L pieces of candidate location data from the M pieces of location data based on the center coordinates of the K interest points, and associating the L pieces of candidate location data with the K interest points; Step S3, for each interest point in the K interest points, calculating the location entropy of the current interest point by taking the candidate location data associated with the current interest point as input, and determining that the current interest point is a non-public place when the location entropy is less than an entropy threshold value; Wherein, the location entropy is calculated by using the following formula: wherein, is the number of times the user u j visited the current point of interest z i , is the total number of times all users in the set of users U visited the current point of interest z i , is the proportion of the number of times the user u j visited the current point of interest z i to the total number of times all users in the set of users U visited the current point of interest z i . 2.A method for determining a user social behavior relationship based on a spatiotemporal co-occurrence region, characterized in that, The method comprises: Step S1, determining a region range of each piece of location data contained in a location data set of a user by using the location data set, and extracting K interest points in the region range; Wherein, the location data set contains M pieces of location data of N users, the location data is coordinate data of a location visited by the user or identification data of a region visited by the user; the interest point is a semantic hot spot position on a map, and each interest point has a center coordinate; Step S2, determining L pieces of candidate location data from the M pieces of location data based on the center coordinates of the K interest points, and associating the L pieces of candidate location data with the K interest points; Step S3, for each interest point in the K interest points, obtaining at least two pieces of candidate location data in a preset time period from the candidate location data associated with the current interest point, and further determining at least two users to which the at least two pieces of candidate location data belong, and determining a region containing the current interest point and the at least two pieces of candidate location data as a space-time co-occurrence region between the at least two users; Step S4, for each interest point in the K interest points, calculating the location entropy of the current interest point by taking the candidate location data associated with the current interest point as input, and determining that the current interest point is a non-public place when the location entropy is less than an entropy threshold value; Wherein, the location entropy is calculated by using the following formula: wherein, is the number of times the user u j visited the current point of interest z i , is the total number of times all users in the set of users U visited the current point of interest z i , is the proportion of the number of times the user u j visited the current point of interest z i to the total number of times all users in the set of users U visited the current point of interest z i . Step S5, extracting the interest points contained in the space-time co-occurrence region and the associated users, and determining a social behavior relationship between the associated users based on the non-public place when the contained interest points are the non-public place.

3. A system for determining non-public places, characterized by The system comprises: A first processing unit configured to determine a region range of each piece of location data contained in a location data set of a user by using the location data set, and extract K interest points in the region range; The position data set contains M pieces of position data of N users, the position data is coordinate data of positions visited by the users or identification data of areas visited by the users; the interest point is a semantic hot position on a map, each interest point has a center coordinate; The second processing unit is configured to determine L pieces of candidate position data from the M pieces of position data based on the center coordinates of the K interest points, and associate the L pieces of candidate position data with the K interest points; The third processing unit is configured to, for each of the K interest points, calculate a position entropy of the current interest point by taking the candidate position data associated with the current interest point as input, and determine that the current interest point is a non-public place when the position entropy is less than an entropy threshold; The position entropy is calculated by using the following formula: wherein, is the number of times the user u j visited the current point of interest z i , is the total number of times all users in the set of users U visited the current point of interest z i , is the proportion of times the user u j visited the current point of interest z i to the total number of times all users in the set of users U visited the current point of interest z i .

4. A system for determining user social behavioral relationships based on spatiotemporal co-occurrence zones, characterized in that, The system comprises: The first processing unit is configured to determine a region range of each piece of position data contained in a position data set of a user by using the position data set, and extract K interest points in the region range; The position data set contains M pieces of position data of N users, the position data is coordinate data of positions visited by the users or identification data of areas visited by the users; the interest point is a semantic hot position on a map, each interest point has a center coordinate; The second processing unit is configured to determine L pieces of candidate position data from the M pieces of position data based on the center coordinates of the K interest points, and associate the L pieces of candidate position data with the K interest points; The third processing unit is configured to, for each of the K interest points, obtain at least two pieces of candidate position data in a preset time period from the candidate position data associated with the current interest point, and further determine at least two users to which the at least two pieces of candidate position data belong, and determine a region containing the current interest point and the at least two pieces of candidate position data as a spatio-temporal co-occurrence region between the at least two users; The fourth processing unit is configured to, for each of the K interest points, calculate a position entropy of the current interest point by taking the candidate position data associated with the current interest point as input, and determine that the current interest point is a non-public place when the position entropy is less than an entropy threshold; The position entropy is calculated by using the following formula: wherein, is the number of times the user u j visits the current point of interest z i , is the total number of times all users in the set of users U visit the current point of interest z i , is the proportion of the number of times the user u j visits the current point of interest z i to the total number of times all users in the set of users U visit the current point of interest z i . The fifth processing unit is configured to extract an interest point contained in the spatio-temporal co-occurrence region and an associated user, and determine a social behavior relationship between the associated users based on the non-public place when the contained interest point is the non-public place.

Images