Point of interest acquisition method, device, electronic device and storage medium

Abstract

The present disclosure is about a method, device, electronic device and storage medium for acquiring points of interest at a resident location, belonging to the field of the Internet, the method comprising: acquiring first point of interest data and first information data, the first point of interest data comprising points of interest corresponding to the information data, the first information data corresponding to the first time information; clustering the first information data to obtain first category information data, the first category information data being the clustering result of the first information data and corresponding to the second time information; determining second category information data corresponding to third time information in the first category information data according to the first category information data and the second time information, the time interval between the third time information and the second time information being a predetermined interval; matching the second information data corresponding to the second category information data in the first point of interest data to obtain second point of interest data. The present disclosure provides a point of interest prediction capability, which supplements and improves the points of interest.

Description

Technical Field

[0001] The present application relates to the field of Internet technology, and in particular to a method, device, electronic device and storage medium for acquiring points of interest. Background Art

[0002] With the advancement of Internet technology and the rise of mobile Internet, the realization of business has a certain correlation with Point of Interest (POI), and more and more technologies are used to obtain POI.

[0003] Existing methods for obtaining points of interest are mainly based on historical data. However, due to the limitations of historical data itself, the accuracy of points of interest is unsatisfactory. Summary of the invention

[0004] The present application provides a method for acquiring points of interest, so as to at least solve the problem of acquiring POIs based on limited information data in the related art.

[0005] According to a first aspect of the embodiment disclosed in the present application, a method for acquiring points of interest is provided, comprising:

[0006] Acquire first point of interest data and first information data, wherein the first point of interest data includes a point of interest corresponding to the information data, and the first information data corresponds to the first time information;

[0007] Clustering the first information data to obtain first category information data, where the first category information data is a clustering result of the first information data and corresponds to the second time information;

[0008] According to the first type of information data and the second time information, determining the second type of information data corresponding to the third time information in the first type of information data, wherein the time interval between the third time information and the second time information is a predetermined interval;

[0009] In the first interest point data, second information data corresponding to the second type of information data is matched to obtain second interest point data.

[0010] As a possible implementation manner, determining the second category information data corresponding to the third time information in the first category information data according to the first category information data and the second time information includes:

[0011] Taking the first type of information data and the second time information as parameters, calculating the first type of information data with the maximum probability of the third time information implied by the first type of information data,

[0012] Determine that the first category information data with the highest probability is the second category information data corresponding to the third time information;

[0013] The time interval of the second time information is greater than the time interval of the third time information.

[0014] As a possible implementation manner, the first category of information data with the maximum probability of the third time information implied by the first category of information data is calculated by taking the first category of information data and the second time information as parameters, including:

[0015] Counting all first point of interest data related to the first type of information data and the third time information to obtain a data set,

[0016] Based on the data set, determine the state transition probability matrix and the observation probability matrix.

[0017] Determine the first type of information data having the maximum probability for each third time information in the data set by using the state transition probability matrix and the observation probability matrix;

[0018] in,

[0019] The state transition probability matrix is ​​used to represent: the probability of the number of times the source information data is transferred to the destination information data between any two first-category information data in the data set occupying the total number of times the source information data is transferred;

[0020] The observation probability matrix is ​​used to represent: the probability that the number of any first-category information data having any third time information in the data set occupies the total number of the first-category information data in the data set.

[0021] As a possible implementation manner, determining the state transition probability matrix includes:

[0022] Construct a state transition probability matrix with M rows and M columns,

[0023] in,

[0024] M is the total number of the first type of information data,

[0025] The matrix element of the i-th row and j-th column is: the ratio of the number of times the i-th first-category information data is transferred to the j-th first-category information data to the total number of times the i-th first-category information data is transferred in the data set.

[0026] As a possible implementation, determining the observation probability matrix includes:

[0027] Construct a state transition probability matrix with M rows and R columns,

[0028] in,

[0029] M is the total number of the first type of information data, R is the total number of the third time information,

[0030] The matrix element of the i-th row and r-th column is: r The ratio of the number of data with the i-th first category information to the total number of information with the i-th first category information data in the data set.

[0031] As a possible implementation manner, the using the state transition probability matrix and the observation probability matrix to determine the first type of information data having the maximum probability for each second time information in the data set includes:

[0032] Using the state transition probability matrix and the observation probability matrix, based on the current previous first-class information data, forward calculate the first-class information data with the maximum probability of the current third-time information,

[0033] When the maximum probability is greater than the set probability threshold, the current last first-category information data is updated to the first-category information data with the maximum probability for the current third-time information, and the updated first-category information data is used to calculate the first-category information data with the maximum probability for the next third-time information;

[0034] Otherwise, the current previous first category information data is kept unchanged, and the current previous first category information data is used to calculate the first category information data with the maximum probability for the next third time information.

[0035] As a possible implementation manner, the first point of interest data is obtained in the following manner:

[0036] Based on the information data, an information sequence is obtained, wherein the information sequence is a sorting result obtained by sorting the information data in chronological order;

[0037] Based on the information sequence, determining trajectory data;

[0038] Remove abnormal information sequences in the trajectory data to obtain filtered trajectory data;

[0039] Based on the filtered trajectory data, identify the target information data,

[0040] The target information data is matched with points of interest to obtain the first point of interest data.

[0041] As a possible implementation manner, the acquiring the information sequence based on the information data includes:

[0042] Converting the location information in the information data into a character string according to the geo-hash;

[0043] Converting the time information in the information data into time information with a set unit level;

[0044] Aggregating the information data having the same character string and the same time information according to the set time aggregation granularity and the set position aggregation granularity;

[0045] The aggregated information data is sorted according to the time sequence to obtain the information sequence.

[0046] As a possible implementation manner, determining trajectory data based on the information sequence includes:

[0047] Traversing each information point data in the information sequence;

[0048] When the time interval between any adjacent first information point and second information point is greater than the set segmentation time interval threshold, the first information point and the second information point are segmented to obtain trajectory data of at least one trajectory line segment.

[0049] in,

[0050] The information sequence after segmentation is composed of sequence segments determined at the segmentation position, each sequence segment corresponds to a trajectory segment, and each information point corresponds to a trajectory point.

[0051] As a possible implementation, removing abnormal information sequences in the trajectory data to obtain filtered trajectory data includes:

[0052] Detecting information point data of speed anomalies in the trajectory data of each trajectory segment;

[0053] From the trajectory segments without speed anomalies, the longest trajectory segment with the most information points is selected, the information sequence is traversed forward with the starting information point of the longest trajectory segment, and the information sequence is traversed backward with the ending information point of the longest trajectory segment, and the information points with speed anomalies are filtered out to obtain the filtered trajectory data.

[0054] As a possible implementation manner, the detecting of information points of speed anomaly in the trajectory data of each trajectory segment includes:

[0055] In the trajectory data of each trajectory segment, information points whose speed between any adjacent information points is greater than a set speed threshold are detected.

[0056] As a possible implementation, identifying target information based on the filtered trajectory data includes:

[0057] Traversing the filtered trajectory data;

[0058] Calculate the distance between the current traversed trajectory point and each subsequent trajectory point until a trajectory point greater than the set distance threshold is found;

[0059] When the time interval between the currently traversed trajectory point and the searched trajectory point is greater than a set time interval threshold, the center point of the trajectory point between the currently traversed trajectory point and the searched trajectory point is used as the target information data, and the next trajectory point adjacent to the searched trajectory point is used as the currently traversed trajectory point;

[0060] In the case where the time interval between the currently traversed trajectory point and the searched trajectory point is not greater than the time interval threshold, the next trajectory point adjacent to the searched trajectory point is used as the currently traversed trajectory point.

[0061] As a possible implementation manner, the first interest point data includes a first polygonal interest point and a first point-shaped interest point, and the performing interest point matching on the target information data to obtain the first interest point data includes:

[0062] Acquire map data in the same area as the target information data, the map data including a second polygonal point of interest and a second point-like point of interest;

[0063] Determine an intersection point between any ray passing through the target information data and the second polygonal interest point, and determine the first polygonal interest point containing the target information data according to a geometric relationship between the intersection point and the target information data;

[0064] Among the second point-like interest points, the first point-like interest point closest to the target information data is determined.

[0065] As a possible implementation manner, matching the second information data corresponding to the second type of information data in the first point of interest data to obtain the second point of interest data includes:

[0066] Mapping first information data corresponding to the second type of information data from the first point of interest data;

[0067] Determine each point of interest matched by the mapped first information data, and merge each point of interest into a union;

[0068] Matching the second type of information data with the union to obtain the second information data,

[0069] Determine that the second information data is the second point of interest data.

[0070] According to a second aspect of the embodiment disclosed in the present application, there is provided an interest point acquisition device, comprising: a first acquisition module, configured to acquire first interest point data and first information data, wherein the first interest point data includes an interest point corresponding to the information data, and the first information data corresponds to first time information;

[0071] a second acquisition module, configured to cluster the first information data to obtain first-category information data, where the first-category information data is a clustering result of the first information data and corresponds to second time information;

[0072] a determination module configured to determine, according to the first type of information data and the second time information, the second type of information data corresponding to the third time information in the first type of information data, wherein the time interval between the third time information and the second time information is a predetermined interval;

[0073] The matching module is configured to match the second information data corresponding to the second type of information data in the first interest point data to obtain the second interest point data.

[0074] As a possible implementation, the determining module is configured to:

[0075] Taking the first type of information data and the second time information as parameters, calculating the first type of information data with the maximum probability of the third time information implied by the first type of information data,

[0076] Determine that the first category information data with the highest probability is the second category information data corresponding to the third time information;

[0077] The time interval of the second time information is greater than the time interval of the third time information.

[0078] As a possible implementation, the determining module is configured to:

[0079] Counting all first point of interest data related to the first type of information data and the third time information to obtain a data set,

[0080] Based on the data set, determine the state transition probability matrix and the observation probability matrix.

[0081] Determine the first type of information data having the maximum probability for each third time information in the data set by using the state transition probability matrix and the observation probability matrix;

[0082] in,

[0083] The state transition probability matrix is ​​used to represent: the probability of the number of times the source information data is transferred to the destination information data between any two first-category information data in the data set occupying the total number of times the source information data is transferred;

[0084] The observation probability matrix is ​​used to represent: the probability that the number of any first-category information data having any third time information in the data set occupies the total number of the first-category information data in the data set.

[0085] As a possible implementation, the determining module is configured to:

[0086] Construct a state transition probability matrix with M rows and M columns,

[0087] in,

[0088] M is the total number of the first type of information data,

[0089] The matrix element of the i-th row and j-th column is: the ratio of the number of times the i-th first-category information data is transferred to the j-th first-category information data to the total number of times the i-th first-category information data is transferred in the data set.

[0090] As a possible implementation, the determining module is configured to:

[0091] Construct a state transition probability matrix with M rows and R columns,

[0092] in,

[0093] M is the total number of the first type of information data, R is the total number of the third time information,

[0094] The matrix element of the i-th row and r-th column is: r The ratio of the number of data with the i-th first category information to the total number of information with the i-th first category information data in the data set.

[0095] As a possible implementation, the determining module is configured to:

[0096] Using the state transition probability matrix and the observation probability matrix, based on the current previous first-class information data, forward calculate the first-class information data with the maximum probability of the current third-time information,

[0097] When the maximum probability is greater than the set probability threshold, the current last first-category information data is updated to the first-category information data with the maximum probability for the current third-time information, and the updated first-category information data is used to calculate the first-category information data with the maximum probability for the next third-time information;

[0098] Otherwise, the current previous first category information data is kept unchanged, and the current previous first category information data is used to calculate the first category information data with the maximum probability for the next third time information.

[0099] As a possible implementation manner, the first acquisition module is configured as follows:

[0100] Based on the information data, an information sequence is obtained, wherein the information sequence is a sorting result obtained by sorting the information data in chronological order;

[0101] Based on the information sequence, determining trajectory data;

[0102] Remove abnormal information sequences in the trajectory data to obtain filtered trajectory data;

[0103] Based on the filtered trajectory data, identify the target information data,

[0104] The target information data is matched with points of interest to obtain the first point of interest data.

[0105] As a possible implementation manner, the first acquisition module is configured as follows:

[0106] Converting the location information in the information data into a character string according to the geo-hash;

[0107] Converting the time information in the information data into time information with a set unit level;

[0108] Aggregating the information data having the same character string and the same time information according to the set time aggregation granularity and the set position aggregation granularity;

[0109] The aggregated information data is sorted according to the time sequence to obtain the information sequence.

[0110] As a possible implementation manner, the first acquisition module is configured as follows:

[0111] Traversing each information point data in the information sequence;

[0112] When the time interval between any adjacent first information point and second information point is greater than the set segmentation time interval threshold, the first information point and the second information point are segmented to obtain trajectory data of at least one trajectory line segment.

[0113] in,

[0114] The information sequence after segmentation is composed of sequence segments determined at the segmentation position, each sequence segment corresponds to a trajectory segment, and each information point corresponds to a trajectory point.

[0115] As a possible implementation manner, the first acquisition module is configured as follows:

[0116] Detecting target information points with abnormal speed in the trajectory data of each trajectory segment;

[0117] From the trajectory segments without speed abnormalities, select the longest trajectory segment with the most target information points, traverse the position sequence information sequence forward with the starting position information point of the longest trajectory segment, and traverse the target information sequence backward with the ending position information point of the longest trajectory segment, filter out the target information points with speed abnormalities, and obtain the filtered trajectory data.

[0118] As a possible implementation manner, the first acquisition module is configured as follows:

[0119] In the trajectory data of each trajectory segment, information points whose speed between any adjacent information points is greater than a set speed threshold are detected.

[0120] As a possible implementation manner, the first acquisition module is configured as follows:

[0121] Traversing the filtered trajectory data;

[0122] Calculate the distance between the current traversed trajectory point and each subsequent trajectory point until a trajectory point greater than the set distance threshold is found;

[0123] When the time interval between the currently traversed trajectory point and the searched trajectory point is greater than a set time interval threshold, the center point of the trajectory point between the currently traversed trajectory point and the searched trajectory point is used as the target information data, and the next trajectory point adjacent to the searched trajectory point is used as the currently traversed trajectory point;

[0124] In the case where the time interval between the currently traversed trajectory point and the searched trajectory point is not greater than the time interval threshold, the next trajectory point adjacent to the searched trajectory point is used as the currently traversed trajectory point.

[0125] As a possible implementation manner, the first interest point data includes a first polygonal interest point and a first point-shaped interest point, and the first acquisition module is configured as follows:

[0126] Acquire map data in the same area as the target information data, the map data including a second polygonal point of interest and a second point-like point of interest;

[0127] Determine an intersection point between any ray passing through the target information data and the second polygonal interest point, and determine the first polygonal interest point containing the target information data according to a geometric relationship between the intersection point and the target information data;

[0128] Among the second point-like interest points, the first point-like interest point closest to the target information data is determined. As a possible implementation, the matching module is configured as follows:

[0129] Mapping first information data corresponding to the second type of information data from the first point of interest data;

[0130] Determine each point of interest matched by the mapped first information data, and merge each point of interest into a union;

[0131] Matching the second type of information data with the union to obtain the second information data,

[0132] Determine that the second information data is the second point of interest data.

[0133] According to a third aspect of the embodiments disclosed in the present application, there is provided an electronic device, including:

[0134] processor;

[0135] a memory for storing executable instructions for the processor;

[0136] The processor is configured to execute the executable instructions to implement any one of the methods for acquiring points of interest.

[0137] According to the fourth aspect of the embodiments disclosed in the present application, a computer-readable storage medium is provided. When at least one instruction in the computer-readable storage medium is executed by a processor of an electronic device, the electronic device can implement a method for acquiring points of interest described in any of the above-mentioned embodiments.

[0138] According to a fifth aspect of the embodiments disclosed in the present application, a computer program product is provided, including a computer program, which implements the above-mentioned method for acquiring points of interest when executed by a processor.

[0139] The technical solutions provided by the embodiments disclosed in this application bring at least the following beneficial effects:

[0140] The point of interest acquisition method provided by the present application captures clustered information data in the first point of interest data by acquiring first point of interest data from information data, thereby providing POI prediction capability when information data is limited, supplementing and improving the first point of interest data, solving the POI acquisition problem based on limited information data, and providing strong support for applications based on information data.

[0141] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0142] The drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the disclosure of the present application, and together with the specification are used to explain the principles disclosed in the present application, and do not constitute an improper limitation on the disclosure of the present application.

[0143] Figure 1 is a flow chart of a POI acquisition method according to an exemplary embodiment;

[0144] Figure 2 A schematic diagram of a process for obtaining first POI data based on information data according to an exemplary embodiment

[0145] Figure 3 is a schematic diagram of obtaining trajectory data according to an exemplary embodiment;

[0146] Figure 4 is a schematic diagram showing identification of target information data according to an exemplary embodiment;

[0147] Figure 5 is a schematic diagram of a process of performing prediction based on first POI data according to an exemplary embodiment;

[0148] Figure 6 is a schematic diagram showing an acquisition state set according to an exemplary embodiment;

[0149] Figure 7 is a flow chart showing forward calculation of first-category information data with maximum probability in each time period according to an exemplary embodiment;

[0150] Figure 8 is a schematic diagram showing the fusion of the second type of information data with the first POI data according to an exemplary embodiment;

[0151] Fig. 9 is a schematic diagram of a device for acquiring points of interest according to an exemplary embodiment;

[0152] Fig.10 The diagram is a schematic diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0153] In order to enable ordinary persons in the art to better understand the technical solutions disclosed in the present application, the technical solutions in the embodiments disclosed in the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0154] It should be noted that the terms "first", "second", etc. in the specification and claims disclosed in the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments disclosed in the present application described here can be implemented in an order other than those illustrated or described here. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the disclosure of the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects disclosed in the present application as detailed in the attached claims.

[0155] The present application proposes a method for acquiring points of interest, which obtains clustered first-category information data and its corresponding second time information from first point of interest data acquired based on information data for prediction, determines second-category information data corresponding to third time information in the first category of information data, and matches the second information data corresponding to the second category of information data in the first point of interest data to obtain final point of interest data.

[0156] See also Figure 1 As shown, Figure 1 FIG. 1 is a flow chart of a method for obtaining POI according to an exemplary embodiment. On the server side, the method includes:

[0157] Step 101, obtaining first point of interest data and first information data, wherein the first point of interest data includes a point of interest corresponding to information data, and the first information data corresponds to first time information;

[0158] As an example, in a location service application scenario, the information data is location information data, the first point of interest data may be a residence location obtained from the location information data, and the point of interest corresponding to the residence location, and the first information data may be location information data with first time information, wherein the first time information may be a set time interval or a set time period.

[0159] Step 102, clustering the first information data to obtain first category information data, where the first category information data is a clustering result of the first information data and corresponds to the second time information;

[0160] Step 103, determining the second type of information data corresponding to the third time information in the first type of information data according to the first type of information data and the second time information, wherein the time interval between the third time information and the second time information is a predetermined interval;

[0161] Step 104: Match the second information data corresponding to the second type of information data in the first point of interest data to obtain second point of interest data.

[0162] This embodiment enriches and improves the first POI data, and effectively combines the information data and the first POI data, which is of great significance for the recommendation of information data. For example, by using POI analysis of location information data, location-based application services are recommended, which is conducive to improving the accuracy of information data push.

[0163] See also Figure 2 As shown, Figure 2 The present invention is a flow chart showing a method of obtaining first POI data based on information data according to an exemplary embodiment. The method comprises:

[0164] Step 201: Based on the information data, an information sequence is obtained, wherein the information sequence is a sorting result obtained by sorting the information data according to time, and each piece of information in the information sequence is each information point.

[0165] For example, if the information data is a collection of geographic location information with a timestamp, the information data will be very dense when the collection frequency is high. As an example, in order to reduce storage and computing costs, the longitude and latitude used to represent the geographic location information are converted into a string, for example, using geohash for conversion; the timestamp is converted into time information with a set unit level, for example, the minute level; then the information data is aggregated according to the set aggregation time granularity and the set aggregation location granularity, so as to aggregate the information data with the same geohash and the same time, and the aggregated information data is sorted in chronological order to obtain an information sequence. In this way, the computing and storage losses are greatly reduced without losing accuracy.

[0166] See also Figure 3 As shown, Figure 3This is a schematic diagram of obtaining trajectory data according to an exemplary embodiment. In the figure, the original information data includes the latitude and longitude of the geographic location information, and timestamp information, wherein the timestamp information includes year, month, day, hour, minute, and second information. The timestamp information is converted into information including year, month, day, hour, and minute, and the longitude and latitude information is converted into a one-dimensional string; the information data with the same timestamp and string are aggregated, for example, according to the aggregation time granularity of minutes and the aggregation location granularity of 9-bit geo-hash, the information data is aggregated, and then sorted in ascending order of time.

[0167] Step 202: Determine trajectory data based on the information sequence.

[0168] In one embodiment, the information sequence is divided according to the time interval between adjacent information points in the information sequence, that is:

[0169] Traverse each information point in the information sequence, and when the time interval between any two adjacent first information points and second information points in the information sequence exceeds the set segmentation time interval threshold, the first information point and the second information point are segmented.

[0170] In this way, all information points that meet the segmentation conditions are segmented, and the information sequence after segmentation is composed of sequence segments determined at the segmentation location, thereby obtaining at least one sequence segment. A sequence segment corresponds to a trajectory segment, and each information point corresponds to a trajectory point in the trajectory segment. In this way, the information data is organized into trajectory data.

[0171] See also Figure 3 As shown in the figure, the information sequence after segmentation forms sequence segments at the segmentation points.

[0172] Step 203 , removing abnormal information in the trajectory data to obtain filtered trajectory data, that is, filtering out abnormal trajectory points in the trajectory segment.

[0173] Given that under normal circumstances, the speed between adjacent track points in a track segment will not exceed the set speed threshold, for example, in the application scenario of location services, the speed threshold may be 140m / s. If the speed between adjacent track points exceeds the set speed threshold, it can be determined as an abnormal track point.

[0174] Based on this, it is possible to detect whether there are information points with abnormal speed in the trajectory data (sequence segments) of each trajectory segment; from the trajectory segments (sequence segments) without abnormal speed, select the sequence segment with the most information points, that is, the longest trajectory segment, traverse each information point in the information sequence forward from the starting information point of the longest trajectory segment (sequence segment), and traverse each information point in the information sequence backward from the ending information point of the longest trajectory segment (sequence segment), filter out the information points with abnormal speed, and obtain the filtered trajectory data.

[0175] See also Figure 3 As shown in the figure, sequence segment B is the sequence segment with the most information points and no information points with abnormal speed. Then, each information point in the information sequence is traversed forward from the starting point in sequence segment B, and each information point in the information sequence is traversed forward from the end information point in sequence segment B to remove the information points with abnormal speed.

[0176] Step 204: Identify target information data based on the filtered trajectory data.

[0177] As an example,

[0178] Traversing the filtered trajectory data;

[0179] Calculate the distance between the current traversed trajectory point and each subsequent trajectory point until a trajectory point greater than the set distance threshold is found;

[0180] When the time interval between the currently traversed trajectory point and the searched trajectory point is greater than a set time interval threshold, the center point of the trajectory point between the currently traversed trajectory point and the searched trajectory point is used as the target information data, and the next trajectory point adjacent to the searched trajectory point is used as the currently traversed trajectory point;

[0181] In the case where the time interval between the currently traversed trajectory point and the searched trajectory point is not greater than the time interval threshold, the next trajectory point adjacent to the searched trajectory point is used as the currently traversed trajectory point.

[0182] For example, in location service application scenarios, target information data containing resident location information can be determined through trajectory data. Figure 4 As shown, Figure 4It is a schematic diagram of identifying target information data according to an exemplary embodiment. Traverse the track points on the filtered track line, such as the current traversal to track point p5, and then calculate the distance between track point p5 and each subsequent track point until a track point whose distance is greater than the set distance threshold is found, such as track point p8. If the time interval between track point p5 and track point p8 is greater than the set dwell time interval threshold, then the center point of the track points included between track point p5 and track point p8 is taken as a dwell position, that is, the first dwell position, and recorded, such as the center point of track points 5, 6, 7, and 8 in the figure (the center of the dotted circle in the figure); then start the traversal from the next track point adjacent to track point p8 as the current traversal point, and repeat it.

[0183] If the time interval between track point p5 and track point p8 is not greater than the set time interval threshold, it means that the residence time condition is not met, and the next track point adjacent to track point p8 is used as the current traversal point, and backward traversal begins.

[0184] This step uses linear complexity target information data recognition to complete the target information data recognition under the time complexity of O(n). Compared with the commonly used spatiotemporal algorithms, the time complexity of target information data recognition is greatly reduced.

[0185] Step 205: Match the target information data with points of interest to obtain first point of interest data.

[0186] As an example, in a location service application scenario, the target information data obtained in step 204 is matched with the POI in the map data.

[0187] Since the data volumes of target information data and POI are both large, in order to reduce time complexity, a preliminary rough match is first performed to obtain the target information data and map data that are in the same area as the target information data, that is, the resident position located in the same area is matched with the POI in the map data.

[0188] Since POIs generally include two categories: one is the second polygonal POI with an area range, which is a set of boundary coordinates, for example, a park with a certain area range, whose POI is a polygon representing the area range of the park; the other is the second point POI, which has only one coordinate point, for example, a small shop or gas station that does not need to represent the area range, so,

[0189] For the second polygonal POI, a ray method is used to find a polygonal POI containing target information data, such as a residence position, specifically: determining an intersection point between any ray passing through the target information data and the polygon, and finding a polygonal interest point containing the target information data according to a geometric relationship between the intersection point and the target information data, thereby obtaining a first polygonal interest point in the first interest point data;

[0190] For the second point-shaped POI, the distance calculation is used to find the Y point-shaped POIs closest to the target information data, and the first point-shaped interest point in the first interest point data is obtained, wherein the size of Y can be set as needed.

[0191] In this way, the target information data and the POI are accurately matched to obtain the first POI data, for example, the POI data of the first resident position. The first POI data is recorded as D1 = {(t1, X1), (t2, X2), ..., (t l ,X l )}, where t l Representation and information data X l The corresponding point of interest, l is the total number of the first POI data, which is a positive integer, and the information data X l Has time attributes.

[0192] In the first POI acquisition process of this embodiment, the large computing consumption caused by the large amount of information data and the O(n2) time complexity of the spatiotemporal algorithm is avoided, and the risk of accuracy loss faced by the calculation after sampling the original information data is avoided.

[0193] See also Figure 5 As shown, Figure 5 This is a flow chart showing a process of making predictions based on first POI data and its associated original information data according to an exemplary embodiment. It includes:

[0194] Step 501 : acquiring a data set for prediction from the first point of interest data obtained in step 205 .

[0195] As an example, it includes:

[0196] Step 5011, acquiring first information data from the first point of interest data according to the set first time information, where the first information data corresponds to the first time information;

[0197] For example, obtain the first POI data within the set first time threshold N and construct a Markov chain:

[0198] Working day: data of working days in the past month + the last information data of the month before + one information data before the forecast day

[0199] Rest day: Data of rest days in the past month + the last information data before the month + one information data before the forecast day

[0200] The first point of interest data is stored locally or on a server, and the data stored on the server is authorized to be accessed.

[0201] Step 5012, clustering the first information data to obtain first category information data, wherein the first category information data is a clustering result of the first information data and corresponds to the second time information;

[0202] See also Figure 6 As shown, Figure 6 It is a schematic diagram of clustering the first information data according to an exemplary embodiment. The first information data within the set second time threshold is clustered to obtain T clustered information data, for example, the information data of historical N days is clustered. The clustering can adopt the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm. Since the clustered data at this time is the information data matching the point of interest in the first point of interest data, the amount of data is much smaller than the original information data, so the use of DBSCAN clustering will not bring too much computing consumption. Wherein, T and N are positive integers.

[0203] From the T cluster information data, the cluster information data with a frequency higher than the set frequency threshold is taken as the state set, and V cluster information data are obtained, which are recorded as {X1,X2,...,X V}, as the state set, V is a positive integer.

[0204] The T cluster information data or the V cluster information data belong to the first category of information data.

[0205] For example, in a location service application, the above clustering process is performed based on the first residence position in the first POI data, so the obtained cluster residence position is the cluster residence position in the first residence position.

[0206] Step 5013, obtaining an observation set;

[0207] According to the third time information, for example, the set time period, it is divided into different observation sets, denoted as {E1, E2, ..., E r}, where E r represents the time period r,

[0208] For example, a day is divided into 24 time periods, and the observation set is recorded as {E1, E2, ..., E 24}.

[0209] The time intervals of the first time information, the second time information and the third time information decrease in sequence.

[0210] Step 5014: Count all first point of interest data related to the first type of information data and the third time information to obtain a data set.

[0211] As an example, a data set for making predictions is obtained based on a state set and an observation set;

[0212] All first POI data related to the state set and the observation set are counted. For example, according to the state set, the first POI data corresponding to the first type of information data is counted. According to the observation set, the first POI data with the third time information is counted. The intersection of the counted POI data is determined as a data set, and a discretized data set is obtained, which is recorded as D2={(t1,X1),(t2,X2),...,(t k ,X k )},t k Indicates the first type of information data X with third time information k points of interest, k is the total number of first POI data in the data set.

[0213] Step 502: determining the second type of information data corresponding to the third time information in the first type of information data according to the first type of information data and the second time information;

[0214] As an example, taking the first category information data and the second time information as parameters, calculating the first category information data with the maximum probability of the third time information implied by the first category information data, and determining that the first category information data with the maximum probability is the second category information data corresponding to the third time information;

[0215] Specifically, based on the data set, prediction is performed to obtain the second type of information data:

[0216] Step 5021, determining a state transition probability matrix, where the state transition probability matrix is ​​used to represent: the probability that the number of times the source information data is transferred to the destination information data between any two first-category information data in the data set occupies the total number of times the source information data is transferred;

[0217] Specifically, a state transition probability matrix with M rows and M columns is constructed,

[0218] in,

[0219] M is the total number of the first type of information data,

[0220] The matrix element of the i-th row and j-th column is: the ratio of the number of times the i-th first-category information data is transferred to the j-th first-category information data to the total number of times the i-th first-category information data is transferred in the data set.

[0221] For example, in a location service application scenario, the state transition probability matrix is ​​used to characterize: between any two resident locations in the data set, the probability of the number of times of leaving from a source resident location of the two resident locations to a destination resident location of the two resident locations occupying the number of times of leaving from the source resident location;

[0222] As an example, a state transition probability matrix P(M×M) with M rows and M columns is constructed, where the element p in the i-th row and j-th column is ij Indicates: The state concentration starts from X i To X j The number of times and the state concentration from X i The ratio of the number of departures; M represents the total number of cluster residence locations. The mathematical formula is:

[0223] p ij =From X i To X j Number of times / from X i The number of times you leave, where i and j are natural numbers and M is a positive integer.

[0224] Step 5022, determining an observation probability matrix, wherein the observation probability matrix is ​​used to characterize: the probability that the number of any first-category information data having any third time information in the data set occupies the total number of the first-category information data in the data set.

[0225] Specifically, a state transition probability matrix with M rows and R columns is constructed.

[0226] in,

[0227] M is the total number of the first type of information data, R is the total number of the third time information,

[0228] The matrix element of the i-th row and r-th column is: r The ratio of the number of data with the first type of information of the i-th type to the total number of data with the first type of information of the i-th type in the data set; r and the third time information E r correspond.

[0229] For example, in a location service application scenario, the observation probability matrix is ​​used to characterize: the number of times in the data set at any set time period at any resident position in the data set, and the probability of occupying the number of times in the data set at the resident position;

[0230] As an example, construct an observation probability matrix O(M×R) with M rows and R columns, where the element o in the i-th row and r-th column is ir Means: Observations are concentrated in time period E r Located in X i The number of observations is concentrated in X i The ratio of the number of times; R is the set time period. The mathematical formula is:

[0231] o ir = In period E r Located in X i The number of observations / observations concentrated in X i The number of

[0232] Among them, i and r are natural numbers, and M and R are positive integers.

[0233] Step 5023, using the state transition probability matrix and the observation probability matrix, determining the first type of information data with the maximum probability for each third time information in the data set;

[0234] As an example, using the state transition probability matrix and the observation probability matrix, based on the current previous first-category information data, the first-category information data with the maximum probability of the current third time information is forward calculated,

[0235] When the maximum probability is greater than the set probability threshold, the current last first-category information data is updated to the first-category information data with the maximum probability for the current third-time information, so as to calculate the first-category information data with the maximum probability for the next third-time information;

[0236] Otherwise, the current previous first category information data is kept unchanged, and the first category information data with the maximum probability of the next third time information is calculated.

[0237] For example, in a location service application scenario, the state transition probability matrix and the observation probability matrix are used to determine the residence position with the maximum probability for each set time period in the data set, and the residence position is the second residence position.

[0238] As an example, using the state transition probability matrix and the observation probability matrix, as well as the last residence position X before , forward calculate the residence position with the maximum probability p in the current time period, when the maximum probability p is greater than the set probability threshold, determine that the residence position of the time period is the residence position with the maximum probability p, and then continue the iterative calculation based on this residence position as the previous state.

[0239] See also Figure 7 As shown, Figure 7This is a flow chart showing a forward calculation of the first type of information data with the maximum probability in each time period according to an exemplary embodiment. The calculation process includes:

[0240] Step 701, initializing the previous first type of information data X before , Time period E r ,

[0241] Step 702: Using the state transition probability matrix and the observation probability matrix, based on the previous first-category information data, forward calculate the first-category information data X with the maximum probability p in the current time period Er. max ,

[0242] Step 703: determine whether the maximum probability p is greater than the set probability threshold. If so, output the first type of information data X with the maximum probability p in the current time period Er. max , the current first-class information data X before Update to time period E r The first type of information data X with the maximum probability p max , update the current time period E r The next adjacent time period E r+1 , return to step 702, the updated first-category information data is used to calculate the first-category information data with the maximum probability in the next time period, and the process is repeated until all time periods in the observation set are calculated.

[0243] Otherwise, keep the current first-category information data unchanged and update the current time period E r The next adjacent time period E r+1 , return to step 702 to calculate the first type of information data with the maximum probability in the next time period, and repeat the process until all time periods in the observation set are calculated.

[0244] In this way, we can obtain the first type of information data with the highest probability in each time period of the observation set, which is used as the predicted second type of information data, denoted as (X max , E r ).

[0245] Step 503: Match the second information data corresponding to the second type of information data in the first point of interest data to obtain second point of interest data.

[0246] As an example, first information data corresponding to the second type of information data is mapped from the first point of interest data;

[0247] Determine each point of interest matched by the mapped first information data, and merge each point of interest into a union;

[0248] Matching the second type of information data with the union to obtain second information data,

[0249] The second information data is determined to be second point of interest data.

[0250] For example, in a location service application scenario, the second resident location is merged with the first resident location to obtain the point of interest data;

[0251] As an example, the second resident position is matched with the union of the POIs of the first resident position corresponding to the second resident position in the POI data of the first resident position;

[0252] As an example, according to the second resident position, the first resident position corresponding to the second resident position is mapped from the POI data of the first resident position; according to the mapped first resident position, the POIs matched thereto are determined, and the matched POIs are merged into a union; the second resident position is matched with the union to obtain the POI data of the second resident position. The matching method may be the same as step 205.

[0253] See also Figure 8 As shown, Figure 8 This is a schematic diagram of the fusion of the second type of information data and the first point of interest according to an exemplary embodiment. Since the second type of information data is clustered, one second type of information data may correspond to multiple target information data, and each of the multiple target information data is matched with a point of interest, and the matched points of interest form a union, which is used for matching the corresponding second type of information data, thereby achieving the effect of supplementing the first POI data.

[0254] The above steps 501 to 503 , based on the predicted POI analysis of the second type of information data, use a hidden Markov model to capture the daily periodicity and regularity in the information data, thereby providing POI acquisition capabilities when the information data is limited.

[0255] See also Fig. 9 As shown, Fig. 9 is a schematic diagram of a device for acquiring a point of interest according to an exemplary embodiment. A first acquisition module 91 is configured to acquire first point of interest data and first information data, wherein the first point of interest data includes a point of interest corresponding to the information data, and the first information data corresponds to the first time information;

[0256] A second acquisition module 92 is configured to cluster the first information data to obtain first-category information data, where the first-category information data is a clustering result of the first information data and corresponds to second time information;

[0257] The determination module 93 is configured to determine the second type of information data corresponding to the third time information in the first type of information data according to the first type of information data and the second time information, wherein the time interval between the third time information and the second time information is a predetermined interval;

[0258] The matching module 94 is configured to match the second information data corresponding to the second type of information data in the first interest point data to obtain the second interest point data.

[0259] The second acquisition module 92 includes:

[0260] The first submodule 921 is configured to obtain first information data from the first point of interest data according to a set first time threshold,

[0261] The second submodule 922 is configured to determine the state set and the observation set,

[0262] The third submodule 923 is configured to count all first point of interest data related to the state set and the observation set to obtain the data set.

[0263] Wherein, the second submodule 922 is configured as follows:

[0264] From the first interest point data, first information data meeting a set second time threshold is screened out, and the screened first information data is clustered to obtain first category information data,

[0265] Filtering out the first category of information data whose frequency is higher than a set frequency threshold from the first category of information data to obtain the state set;

[0266] Determine the observation set according to the set time period.

[0267] The determination module 93 includes:

[0268] The state transition probability matrix calculation submodule 931 is configured to determine the state transition probability matrix.

[0269] The observation probability matrix calculation submodule 932 is configured to determine the observation probability matrix,

[0270] The second identification submodule 933 uses the state transition probability matrix and the observation probability matrix to determine the second type of information data with the maximum probability in each set time period in the data set.

[0271] The state transition probability matrix submodule 931 is configured as follows:

[0272] Construct a state transition probability matrix with M rows and M columns,

[0273] in,

[0274] M is the total number of the first type of information data in the state set,

[0275] The matrix element of the i-th row and j-th column is: the ratio of the number of times the i-th first-category information data is transferred to the j-th first-category information data to the total number of times the i-th first-category information data is transferred in the data set.

[0276] The observation probability matrix submodule 932 is configured as follows:

[0277] Construct a state transition probability matrix with M rows and R columns,

[0278] in,

[0279] M is the total number of the first type of information data in the state set, R is the total number of the set time period,

[0280] The matrix element in row i and column r is: r The ratio of the number of data with the i-th first category information to the total number of data with the i-th first category information in the data set.

[0281] The second identification submodule 933 is configured:

[0282] Using the state transition probability matrix and the observation probability matrix, based on the current previous first-class information data, forward calculate the first-class information data with the maximum probability in the current set time period,

[0283] When the maximum probability is greater than the set probability threshold, the current previous first-category information data is updated to the first-category information data with the maximum probability in the current time period, and the updated first-category information data is used to calculate the first-category information data with the maximum probability in the next time period;

[0284] Otherwise, the current previous first category information data is kept unchanged, and the current previous first category information data is used to calculate the first category information data with the maximum probability in the next time period.

[0285] The matching module 94 includes:

[0286] A mapping submodule 941 is configured to map the first type of information data corresponding to the second type of information data from the first point of interest data;

[0287] The POI union submodule 942 is configured to determine each point of interest matched thereto according to the mapped first type of information data, and merge each point of interest into a union;

[0288] The second interest point matching submodule 943 is configured to match the second type of information data with the union to obtain the second information data, and determine that the second information data is the second interest point data.

[0289] The first acquisition module 91 includes:

[0290] The information sequence acquisition submodule 911 is configured to acquire an information sequence based on the information data, wherein the information sequence is a sorting result obtained by sorting the information data in chronological order;

[0291] A trajectory segmentation submodule 912 is configured to determine trajectory data based on the information sequence;

[0292] A filtering submodule 913 is configured to remove abnormal information data from the trajectory data to obtain filtered trajectory data;

[0293] The first identification submodule 914 is configured to identify target information data based on the filtered trajectory data.

[0294] The first interest point matching submodule 915 is configured to perform interest point matching on the target information data to obtain the first interest point data.

[0295] The information sequence acquisition submodule 911 includes:

[0296] The location information conversion unit 9111 is configured to convert the location information in the information data into a character string according to the geo-hash;

[0297] A time information conversion unit 9112, configured to convert the time information in the information data into time information with a set unit level;

[0298] Aggregation unit 9113, aggregates the information data having the same character string and the same time information according to a set time aggregation granularity and a set position aggregation granularity;

[0299] The sorting unit 9114 sorts the aggregated information data according to the time sequence to obtain the information sequence.

[0300] The trajectory segmentation submodule 912 is configured as follows:

[0301] Traversing each information point in the information sequence;

[0302] When the time interval between any adjacent first information point and second information point is greater than the set segmentation time interval threshold, the first information point and the second information point are segmented to obtain trajectory data of at least one trajectory line segment.

[0303] in,

[0304] The information sequence after segmentation is composed of sequence segments determined at the segmentation position, and each sequence segment corresponds to a trajectory segment.

[0305] The filtering submodule 913 is configured as follows:

[0306] Detect whether there are information points with abnormal speed in the trajectory data of each trajectory segment;

[0307] From the trajectory segments without speed anomalies, the longest trajectory segment with the most information points is selected, the information sequence is traversed forward with the starting information point of the longest trajectory segment, and the information sequence is traversed backward with the ending information point of the longest trajectory segment, and the information points with speed anomalies are filtered out to obtain the filtered trajectory data.

[0308] The filtering submodule is configured 913 as follows:

[0309] In the trajectory data of each trajectory segment, it is detected whether the speed between any adjacent information points is greater than a set speed threshold. If the speed is greater than the speed threshold, it is determined that there is an information point with abnormal speed.

[0310] The first identification submodule 914 is configured to:

[0311] Traversing the filtered trajectory data;

[0312] Calculate the distance between the currently traversed information point and the subsequent information point until an information point with a distance greater than the set distance threshold is found;

[0313] When the time interval between the currently traversed information point and the searched information point is greater than the set time interval threshold, the center point of the information points between the currently traversed information point and the searched information point is used as the target information data, and the next information point adjacent to the searched information point is used as the currently traversed information point;

[0314] In the case that the time interval between the currently traversed information point and the searched information point is not greater than the time interval threshold, the next information point adjacent to the searched information point is used as the currently traversed information point.

[0315] The first interest point matching submodule 915 is configured as follows:

[0316] Acquire map data in the same area as the target information data, the map data including a second polygonal point of interest and a second point-like point of interest;

[0317] Determine an intersection point between any ray passing through the target information data and the second polygonal interest point, and determine the first polygonal interest point containing the target information data according to a geometric relationship between the intersection point and the target information data;

[0318] Among the second point-like interest points, the first point-like interest point closest to the target information data is determined.

[0319] See also Fig.10 As shown, Fig.10 The electronic device includes a processor; a memory for storing executable instructions of the processor;

[0320] The processor is configured to execute the executable instructions to implement any one of the methods for acquiring points of interest.

[0321] In an exemplary embodiment, a computer-readable storage medium including at least one instruction is also provided, such as a memory including at least one instruction. The at least one instruction can be executed by a processor in a computer device to complete a method for acquiring points of interest described in the above embodiment.

[0322] Optionally, the above-mentioned computer-readable storage medium may be a non-temporary computer-readable storage medium, for example, the non-temporary computer-readable storage medium may include ROM (Read-Only Memory), RAM (Random-Access Memory), CD-ROM (Compact Disc Read-Only Memory), magnetic tape, floppy disk and optical data storage device, etc.

[0323] In an exemplary embodiment, a computer program product is also provided, including one or more instructions, and the one or more instructions can be executed by a processor of a computer device to complete the business tag display method described in each of the above embodiments.

[0324] Those skilled in the art will readily appreciate other embodiments disclosed in the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation disclosed in the present application, which follows the general principles disclosed in the present application and includes common knowledge or customary technical means in the art that are not disclosed in the present application. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

[0325] It should be understood that the present disclosure is not limited to the precise structures described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for acquiring points of interest, characterized in that: include: Acquire first point of interest data and first information data, wherein the first point of interest data includes a point of interest corresponding to the information data, and the first information data corresponds to the first time information; Clustering the first information data to obtain first category information data, where the first category information data is a clustering result of the first information data and corresponds to the second time information; According to the first type of information data and the second time information, determining the second type of information data corresponding to the third time information in the first type of information data, wherein the time interval of the third time information and the time interval of the second time information are predetermined intervals; Matching the second information data corresponding to the second type of information data in the first point of interest data to obtain second point of interest data; in, The determining, according to the first category of information data and the second time information, the second category of information data corresponding to the third time information in the first category of information data comprises: Taking the first type of information data and the second time information as parameters, calculating the first type of information data with the maximum probability of the third time information implied by the first type of information data, It is determined that the first type of information data with the maximum probability is the second type of information data corresponding to the third time information.

2. The method according to claim 1, characterized in that: The time interval of the second time information is greater than the time interval of the third time information.

3. The method according to claim 2, characterized in that The method of calculating the first type of information data with the maximum probability of the third time information implied by the first type of information data by taking the first type of information data and the second time information as parameters includes: Counting all first point of interest data related to the first type of information data and the third time information to obtain a data set, Based on the data set, determine the state transition probability matrix and the observation probability matrix. Determine the first type of information data having the maximum probability for each third time information in the data set by using the state transition probability matrix and the observation probability matrix; in, The state transition probability matrix is ​​used to represent: the probability of the number of times the source information data is transferred to the destination information data between any two first-category information data in the data set occupying the total number of times the source information data is transferred; The observation probability matrix is ​​used to represent: the probability that the number of any first-category information data having any third time information in the data set occupies the total number of the first-category information data in the data set.

4. The method according to claim 3, characterized in that The determining of the state transition probability matrix comprises: Construct a state transition probability matrix with M rows and M columns, in, M is the total number of the first type of information data, The matrix element of the i-th row and j-th column is: the ratio of the number of times the i-th first-category information data is transferred to the j-th first-category information data to the total number of times the i-th first-category information data is transferred in the data set.

5. The method according to claim 3, characterized in that: The observation probability matrix is ​​determined as follows: Construct a state transition probability matrix with M rows and R columns, in, M is the total number of the first type of information data, R is the total number of the third time information, The matrix element of the i-th row and r-th column is: r The ratio of the number of data with the i-th first category information to the total number of information with the i-th first category information data in the data set.

6. The method according to claim 3, characterized in that The determining, using the state transition probability matrix and the observation probability matrix, that each second time information in the data set has the first type of information data with the maximum probability comprises: Using the state transition probability matrix and the observation probability matrix, based on the current previous first-class information data, forward calculate the first-class information data with the maximum probability of the current third-time information, When the maximum probability is greater than the set probability threshold, the current last first-category information data is updated to the first-category information data with the maximum probability for the current third-time information, and the updated first-category information data is used to calculate the first-category information data with the maximum probability for the next third-time information; Otherwise, the current previous first category information data is kept unchanged, and the current previous first category information data is used to calculate the first category information data with the maximum probability for the next third time information.

7. The method according to claim 3, characterized in that The first point of interest data is obtained in the following manner: Based on the information data, an information sequence is obtained, wherein the information sequence is a sorting result obtained by sorting the information data in chronological order; Based on the information sequence, determining trajectory data; Remove abnormal information sequences in the trajectory data to obtain filtered trajectory data; Based on the filtered trajectory data, target information data is identified, wherein the target information data is used to characterize: the center point of the trajectory points included in the case where the distance between the trajectory points is greater than a set distance threshold and the time interval between the trajectory points is greater than a set time interval threshold; The target information data is matched with points of interest to obtain the first point of interest data.

8. The method according to claim 7, characterized in that The step of acquiring an information sequence based on the information data comprises: Converting the location information in the information data into a character string according to the geo-hash; Converting the time information in the information data into time information with a set unit level; Aggregating the information data having the same character string and the same time information according to the set time aggregation granularity and the set position aggregation granularity; The aggregated information data is sorted according to the time sequence to obtain the information sequence.

9. The method according to claim 7, characterized in that: The determining of trajectory data based on the information sequence includes: Traversing each information point data in the information sequence; When the time interval between any adjacent first information point and second information point is greater than the set segmentation time interval threshold, the first information point and the second information point are segmented to obtain trajectory data of at least one trajectory line segment. in, The information sequence after segmentation is composed of sequence segments determined at the segmentation position, each sequence segment corresponds to a trajectory segment, and each information point corresponds to a trajectory point.

10. The method according to claim 9, characterized in that The step of removing abnormal information sequences from the trajectory data to obtain filtered trajectory data includes: Detecting information point data of speed anomalies in the trajectory data of each trajectory segment; From the trajectory segments without speed anomalies, the longest trajectory segment with the most information points is selected, the information sequence is traversed forward with the starting information point of the longest trajectory segment, and the information sequence is traversed backward with the ending information point of the longest trajectory segment, and the information points with speed anomalies are filtered out to obtain the filtered trajectory data.

11. The method according to claim 10, characterized in that The detecting of information points of speed anomaly existing in the trajectory data of each trajectory segment includes: In the trajectory data of each trajectory segment, information points whose speed between any adjacent information points is greater than a set speed threshold are detected.

12. The method according to claim 7, characterized in that The identifying target information data based on the filtered trajectory data includes: Traversing the filtered trajectory data; Calculate the distance between the current traversed trajectory point and each subsequent trajectory point until a trajectory point greater than the set distance threshold is found; When the time interval between the currently traversed trajectory point and the searched trajectory point is greater than a set time interval threshold, the center point of the trajectory point between the currently traversed trajectory point and the searched trajectory point is used as the target information data, and the next trajectory point adjacent to the searched trajectory point is used as the currently traversed trajectory point; In the case where the time interval between the currently traversed trajectory point and the searched trajectory point is not greater than the time interval threshold, the next trajectory point adjacent to the searched trajectory point is used as the currently traversed trajectory point.

13. The method according to claim 7, characterized in that The first interest point data includes a first polygonal interest point and a first point-shaped interest point, and performing interest point matching on the target information data to obtain the first interest point data includes: Acquire map data in the same area as the target information data, the map data including a second polygonal point of interest and a second point-like point of interest; Determine an intersection point between any ray passing through the target information data and the second polygonal interest point, and determine the first polygonal interest point containing the target information data according to a geometric relationship between the intersection point and the target information data; Among the second point-like interest points, the first point-like interest point closest to the target information data is determined.

14. The method according to claim 1, characterized in that The step of matching the second information data corresponding to the second type of information data in the first point of interest data to obtain the second point of interest data includes: Mapping first information data corresponding to the second type of information data from the first point of interest data; Determine each point of interest matched by the mapped first information data, and merge each point of interest into a union; Matching the second type of information data with the union to obtain the second information data, Determine that the second information data is the second point of interest data.

15. A device for acquiring points of interest, characterized in that: include: A first acquisition module is configured to acquire first interest point data and first information data, wherein the first interest point data includes an interest point corresponding to the information data, and the first information data corresponds to the first time information; a second acquisition module, configured to cluster the first information data to obtain first-category information data, where the first-category information data is a clustering result of the first information data and corresponds to second time information; a determination module configured to determine, according to the first type of information data and the second time information, the second type of information data corresponding to the third time information in the first type of information data, wherein the time interval of the third time information and the time interval of the second time information are a predetermined interval; a matching module, configured to match the second information data corresponding to the second type of information data in the first point of interest data to obtain second point of interest data; in, The determining, according to the first category of information data and the second time information, the second category of information data corresponding to the third time information in the first category of information data comprises: Taking the first type of information data and the second time information as parameters, calculating the first type of information data with the maximum probability of the third time information implied by the first type of information data, It is determined that the first type of information data with the maximum probability is the second type of information data corresponding to the third time information.

16. The device according to claim 15, characterized in that The time interval of the second time information is greater than the time interval of the third time information.

17. The device according to claim 16, characterized in that The determination module is configured to: Counting all first point of interest data related to the first type of information data and the third time information to obtain a data set, Based on the data set, determine the state transition probability matrix and the observation probability matrix. Determine the first type of information data having the maximum probability for each third time information in the data set by using the state transition probability matrix and the observation probability matrix; in, The state transition probability matrix is ​​used to represent: the probability of the number of times the source information data is transferred to the destination information data between any two first-category information data in the data set occupying the total number of times the source information data is transferred; The observation probability matrix is ​​used to represent: the probability that the number of any first-category information data having any third time information in the data set occupies the total number of the first-category information data in the data set.

18. The device according to claim 17, characterized in that The determination module is configured to: Construct a state transition probability matrix with M rows and M columns, in, M is the total number of the first type of information data, The matrix element of the i-th row and j-th column is: the ratio of the number of times the i-th first-category information data is transferred to the j-th first-category information data to the total number of times the i-th first-category information data is transferred in the data set.

19. The device according to claim 17, characterized in that The determination module is configured to: Construct a state transition probability matrix with M rows and R columns, in, M is the total number of the first type of information data, R is the total number of the third time information, The matrix element of the i-th row and r-th column is: r The ratio of the number of data with the i-th first category information to the total number of information with the i-th first category information data in the data set.

20. The device according to claim 17, characterized in that The determination module is configured to: Using the state transition probability matrix and the observation probability matrix, based on the current previous first-class information data, forward calculate the first-class information data with the maximum probability of the current third-time information, When the maximum probability is greater than the set probability threshold, the current last first-category information data is updated to the first-category information data with the maximum probability for the current third-time information, and the updated first-category information data is used to calculate the first-category information data with the maximum probability for the next third-time information; Otherwise, the current previous first category information data is kept unchanged, and the current previous first category information data is used to calculate the first category information data with the maximum probability for the next third time information.

21. The device according to claim 17, characterized in that The first acquisition module is configured as follows: Based on the information data, an information sequence is obtained, wherein the information sequence is a sorting result obtained by sorting the information data in chronological order; Based on the information sequence, determining trajectory data; Remove abnormal information sequences in the trajectory data to obtain filtered trajectory data; Based on the filtered trajectory data, target information data is identified, wherein the target information data is used to characterize: the center point of the trajectory points included in the case where the distance between the trajectory points is greater than a set distance threshold and the time interval between the trajectory points is greater than a set time interval threshold; The target information data is matched with points of interest to obtain the first point of interest data.

22. The device according to claim 21, characterized in that The first acquisition module is configured as follows: Converting the location information in the information data into a character string according to the geo-hash; Converting the time information in the information data into time information with a set unit level; Aggregating the information data having the same character string and the same time information according to the set time aggregation granularity and the set position aggregation granularity; The aggregated information data is sorted according to the time sequence to obtain the information sequence.

23. The device according to claim 21, characterized in that The first acquisition module is configured as follows: Traversing each information point data in the information sequence; When the time interval between any adjacent first information point and second information point is greater than the set segmentation time interval threshold, the first information point and the second information point are segmented to obtain trajectory data of at least one trajectory line segment. in, The information sequence after segmentation is composed of sequence segments determined at the segmentation position, each sequence segment corresponds to a trajectory segment, and each information point corresponds to a trajectory point.

24. The device according to claim 23, characterized in that The first acquisition module is configured as follows: Detecting target information points with abnormal speed in the trajectory data of each trajectory segment; From the trajectory segments without speed abnormalities, select the longest trajectory segment with the most target information points, traverse the information sequence forward with the starting position information point of the longest trajectory segment, and traverse the information sequence backward with the ending position information point of the longest trajectory segment, filter out the target information points with speed abnormalities, and obtain the filtered trajectory data.

25. The device according to claim 23, characterized in that The first acquisition module is configured as follows: In the trajectory data of each trajectory segment, information points whose speed between any adjacent information points is greater than a set speed threshold are detected.

26. The device according to claim 21, characterized in that The first acquisition module is configured as follows: Traversing the filtered trajectory data; Calculate the distance between the current traversed trajectory point and each subsequent trajectory point until a trajectory point greater than the set distance threshold is found; When the time interval between the currently traversed trajectory point and the searched trajectory point is greater than a set time interval threshold, the center point of the trajectory point between the currently traversed trajectory point and the searched trajectory point is used as the target information data, and the next trajectory point adjacent to the searched trajectory point is used as the currently traversed trajectory point; In the case where the time interval between the currently traversed trajectory point and the searched trajectory point is not greater than the time interval threshold, the next trajectory point adjacent to the searched trajectory point is used as the currently traversed trajectory point.

27. The device according to claim 21, characterized in that The first interest point data includes a first polygonal interest point and a first point-shaped interest point, and the first acquisition module is configured to: Acquire map data in the same area as the target information data, the map data including a second polygonal point of interest and a second point-like point of interest; Determine an intersection point between any ray passing through the target information data and the second polygonal interest point, and determine the first polygonal interest point containing the target information data according to a geometric relationship between the intersection point and the target information data; Among the second point-like interest points, the first point-like interest point closest to the target information data is determined.

28. The device according to claim 15, characterized in that The matching module is configured as follows: Mapping first information data corresponding to the second type of information data from the first point of interest data; Determine each point of interest matched by the mapped first information data, and merge each point of interest into a union; Matching the second type of information data with the union to obtain the second information data, Determine that the second information data is the second point of interest data.

29. An electronic device, characterized in that: include: processor; a memory for storing executable instructions for the processor; The processor is configured to execute the executable instructions to implement a method for acquiring points of interest as described in any one of claims 1 to 14.

30. A computer-readable storage medium, characterized in that: When at least one instruction in the computer-readable storage medium is executed by a processor of an electronic device, the electronic device can implement a method for acquiring points of interest as described in any one of claims 1 to 14.

31. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for acquiring points of interest as claimed in any one of claims 1 to 14 is implemented.

Images