Point of interest recall method, apparatus, electronic device, and storage medium

Abstract

The application discloses a point-of-interest recall method and device, electronic equipment and a storage medium. The method comprises the following steps: obtaining a target point-of-interest to be processed; determining a target recall distance of the target point-of-interest according to a target point-of-interest type of the target point-of-interest and a recall point-of-interest type corresponding to the target point-of-interest type; determining a recall position range of the target point-of-interest according to the target recall distance and position information of the target point-of-interest; searching for a recall point-of-interest corresponding to the target point-of-interest in the recall position range, and recalling the recall point-of-interest. The method obtains the target recall distance through the target point-of-interest type and the recall point-of-interest type, and constructs the recall position range. Therefore, while limiting the search range when searching for the recall point-of-interest, the accuracy of the recall position range is improved. The number of point-of-interests that need to be searched when searching for the recall point-of-interest is reduced, the efficiency of recall is improved, the accuracy of the recall point-of-interest is ensured, and missing selection and wrong selection are avoided.

Description

Technical Field

[0001] This application relates to the field of electronic map technology, specifically to a method, apparatus, electronic device, and storage medium for recalling points of interest. Background Technology

[0002] With the rapid development of the internet, duplicate names are common for POIs (Points of Interest) in the input and use of location information. Furthermore, due to the time lag between data updates and the real world, and the possibility of POIs being renamed, relocated, closed, or newly established, these factors all impact the accurate use of POI data. The keyword search method commonly used in the market can, to some extent, avoid the problems caused by duplicate information.

[0003] However, the current POI recall methods have limitations in accuracy and cannot meet the increasingly sophisticated application needs of POIs. Summary of the Invention

[0004] This application provides a method, apparatus, electronic device, and storage medium for recalling points of interest, aiming to solve the problem of inaccuracy in current point of interest recall methods.

[0005] Firstly, this application provides a method for recalling points of interest, including:

[0006] Obtain the target interest points to be processed;

[0007] Based on the target interest point type and the corresponding recall interest point type, determine the target recall distance of the target interest point.

[0008] Based on the target recall distance and the location information of the target point of interest, the recall location range of the target point of interest is determined;

[0009] Within the recall location range, the recall point of interest corresponding to the target point of interest is obtained, and the recall point of interest is recalled.

[0010] In one possible implementation of this application, determining the target recall distance of the target interest point based on its target interest point type and the corresponding recall interest point type includes:

[0011] Obtain the target interest point type, the recall interest point type corresponding to the target interest point type, the first word vector corresponding to the target interest point type, and the second word vector corresponding to the recall interest point type;

[0012] Based on the target interest point type and the corresponding recall interest point type, a recall distance prediction model is selected from the preset candidate models.

[0013] The first word vector and the second word vector are fused and input into the recall distance prediction model to predict the target recall distance of the target interest point.

[0014] In one possible implementation of this application, the target recall distance has at least two values.

[0015] The step of determining the recall location range of the target interest point based on the target recall distance and the location information of the target interest point includes:

[0016] If the target interest point includes a target subject word, then the recall location range of the target interest point is determined based on the minimum target recall distance and the location information of the target interest point, wherein the target subject word is a subject word in a preset chain brand database;

[0017] If the target interest point does not include the target subject word, then the recall location range of the target interest point is determined based on the maximum target recall distance and the location information of the target interest point.

[0018] In one possible implementation of this application, before determining the target recall distance of the target interest point based on the target interest point type and the recall interest point type corresponding to the target interest point type, the method further includes:

[0019] The subject words in the target interest points are compared with the subject words in a preset first database, wherein the first database contains the interest point types corresponding to the preset first subject words;

[0020] If the subject word in the target interest point exists in the first database, then the interest point type corresponding to the subject word in the first database is set as the target interest point type of the target interest point;

[0021] If the subject word in the target interest point does not exist in the first database, then based on the subject word and its word position, the interest point type corresponding to the subject word is obtained from a preset second database, and the interest point type is set as the target interest point type of the target interest point. The word position includes one of the beginning, middle, and end of the word, and the second database contains preset interest point types corresponding to the second subject word in different word positions.

[0022] In one possible implementation of this application, the recall location range includes a first location range and a second location range, wherein the first location range is rectangular, and the area of ​​the second location range is smaller than the area of ​​the first location range.

[0023] The step of searching for and retrieving the recall point of interest corresponding to the target point of interest within the recall location range, and recalling the recall point of interest, includes:

[0024] Obtain the first candidate point of interest whose location is within the range of the first location;

[0025] The similarity between the first candidate interest point and the target interest point is compared with a preset similarity threshold to obtain a second candidate interest point with a similarity greater than the similarity threshold.

[0026] The second candidate point of interest whose location is within the second location range is set as the recall point of interest corresponding to the target point of interest;

[0027] The recall refers to the points of interest mentioned above.

[0028] In one possible implementation of this application, obtaining the first candidate point of interest whose location is within the range of the first location includes:

[0029] The initial points of interest located within the recall location range are used as points of interest to be filtered, and the recording time of the points of interest to be filtered is obtained.

[0030] The points of interest to be screened are filtered based on the recording duration to obtain the first candidate points of interest with a recording duration greater than or equal to a preset time threshold.

[0031] In one possible implementation of this application, obtaining the target interest point to be processed includes:

[0032] Receive an input instruction for points of interest and obtain the newly added points of interest corresponding to the input instruction;

[0033] The newly added point of interest is set as the target point of interest to be processed.

[0034] Secondly, this application provides a point-of-interest recall device, comprising:

[0035] The acquisition unit is used to acquire the target interest points to be processed.

[0036] The distance determination unit is used to determine the target recall distance of the target interest point based on the target interest point type and the recall interest point type corresponding to the target interest point type.

[0037] The range determination unit is used to determine the recall location range of the target interest point based on the target recall distance and the location information of the target interest point;

[0038] The recall unit is used to search for and retrieve the recall point of interest corresponding to the target point of interest within the recall location range.

[0039] In one possible implementation of this application, the distance determining unit is further used for:

[0040] Obtain the target interest point type, the recall interest point type corresponding to the target interest point type, the first word vector corresponding to the target interest point type, and the second word vector corresponding to the recall interest point type;

[0041] Based on the target interest point type and the corresponding recall interest point type, a recall distance prediction model is selected from the preset candidate models.

[0042] The first word vector and the second word vector are fused and input into the recall distance prediction model to predict the target recall distance of the target interest point.

[0043] In one possible implementation of this application, the range determination unit is further configured to:

[0044] The step of determining the recall location range of the target interest point based on the target recall distance and the location information of the target interest point includes:

[0045] If the target interest point includes a target subject word, then the recall location range of the target interest point is determined based on the minimum target recall distance and the location information of the target interest point, wherein the target subject word is a subject word in a preset chain brand database;

[0046] If the target interest point does not include the target subject word, then the recall location range of the target interest point is determined based on the maximum target recall distance and the location information of the target interest point.

[0047] In one possible implementation of this application, the distance determining unit is further used for:

[0048] The subject words in the target interest points are compared with the subject words in a preset first database, wherein the first database contains the interest point types corresponding to the preset first subject words;

[0049] If the subject word in the target interest point exists in the first database, then the interest point type corresponding to the subject word in the first database is set as the target interest point type of the target interest point;

[0050] If the subject word in the target interest point does not exist in the first database, then based on the subject word and its word position, the interest point type corresponding to the subject word is obtained from a preset second database, and the interest point type is set as the target interest point type of the target interest point. The word position includes one of the beginning, middle, and end of the word, and the second database contains preset interest point types corresponding to the second subject word in different word positions.

[0051] In one possible implementation of this application, the recall unit is further configured to:

[0052] Obtain the first candidate point of interest whose location is within the range of the first location;

[0053] The similarity between the first candidate interest point and the target interest point is compared with a preset similarity threshold to obtain a second candidate interest point with a similarity greater than the similarity threshold.

[0054] The second candidate point of interest whose location is within the second location range is set as the recall point of interest corresponding to the target point of interest;

[0055] The recall refers to the points of interest mentioned above.

[0056] In one possible implementation of this application, the recall unit is further configured to:

[0057] The initial points of interest located within the recall location range are used as points of interest to be filtered, and the recording time of the points of interest to be filtered is obtained.

[0058] The points of interest to be screened are filtered based on the recording duration to obtain the first candidate points of interest with a recording duration greater than or equal to a preset time threshold.

[0059] The initial points of interest located within the recall location range are used as points of interest to be filtered, and the entry time of the points of interest to be filtered is obtained.

[0060] The points of interest to be screened are filtered based on the entry time to obtain the first candidate points of interest.

[0061] In one possible implementation of this application, the acquiring unit is further configured to:

[0062] Receive an input instruction for points of interest and obtain the newly added points of interest corresponding to the input instruction;

[0063] The newly added point of interest is set as the target point of interest to be processed.

[0064] Thirdly, this application also provides an electronic device, which includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor invokes the computer program in the memory, it executes the steps in any of the point of interest recall methods provided in this application.

[0065] Fourthly, this application also provides a storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the point-of-interest recall methods provided in this application.

[0066] In summary, the point of interest recall method provided in this application includes: acquiring a target point of interest to be processed; determining a target recall distance for the target point of interest based on the target point of interest type and the recall point of interest type corresponding to the target point of interest type; determining a recall location range for the target point of interest based on the target recall distance and the location information of the target point of interest; searching for and retrieving the recall point of interest corresponding to the target point of interest within the recall location range; and recalling the recall point of interest.

[0067] As can be seen, the point of interest recall method provided in this application obtains the target recall distance by using the target point of interest type and the recall point of interest type, and then constructs the recall location range based on the target recall distance. Therefore, while limiting the search range when searching for recall points of interest, it improves the accuracy of the recall location range. This reduces the number of points of interest that need to be searched when searching for recall points of interest, reduces the computational load of point of interest recall, improves the efficiency of point of interest recall, and ensures the accuracy of obtaining recall points of interest, avoiding missed selections and incorrect selections. Attached Figure Description

[0068] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0069] Figure 1 This is a schematic diagram illustrating an application scenario of the point-of-interest recall method provided in the embodiments of this application;

[0070] Figure 2 This is a flowchart illustrating an interest point recall method provided in an embodiment of this application;

[0071] Figure 3 This is a schematic diagram of a process for obtaining the target recall distance provided in an embodiment of this application;

[0072] Figure 4This is a flowchart illustrating a recall process for points of interest obtained through recall, as provided in an embodiment of this application.

[0073] Figure 5 This is a schematic diagram of an embodiment of the point-of-interest recall device provided in this application.

[0074] Figure 6 This is a schematic diagram of an embodiment of the electronic device provided in this application. Detailed Implementation

[0075] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0076] In the description of the embodiments of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0077] To enable any person skilled in the art to implement and use this application, the following description is provided. In this description, details are set forth for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be implemented without using these specific details. In other instances, well-known processes will not be described in detail to avoid obscuring the description of the embodiments of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in the embodiments of this application.

[0078] This application provides a method, apparatus, electronic device, and storage medium for recalling points of interest. The point-of-interest recall apparatus can be integrated into an electronic device, which may be a server or a terminal, etc.

[0079] The execution subject of the point of interest recall method in this application embodiment can be the point of interest recall device provided in this application embodiment, or different types of electronic devices such as server equipment, physical host, or user equipment (UE) that integrate the point of interest recall device. The point of interest recall device can be implemented in hardware or software. The UE can be a terminal device such as a smartphone, tablet computer, laptop computer, handheld computer, desktop computer, or personal digital assistant (PDA).

[0080] The electronic device can operate independently or in a cluster.

[0081] See Figure 1 , Figure 1 This is a schematic diagram of a point-of-interest (POI) recall system provided in an embodiment of this application. The POI recall system may include an electronic device 101, which integrates a POI recall device.

[0082] In addition, such as Figure 1 As shown, the point-of-interest recall system may also include a memory 102 for storing data, such as text data.

[0083] It should be noted that, Figure 1 The schematic diagram of the point-of-interest (POI) recall system shown is merely an example. The POI recall system and scenarios described in this application are intended to more clearly illustrate the technical solutions of this application and do not constitute a limitation on the technical solutions provided in this application. As those skilled in the art will know, with the evolution of POI recall systems and the emergence of new business scenarios, the technical solutions provided in this invention are also applicable to similar technical problems.

[0084] The following describes the point-of-interest (POI) recall method provided in this application embodiment. In this application embodiment, an electronic device is used as the execution subject. For simplicity and ease of description, the execution subject will be omitted in subsequent method embodiments. The POI recall method includes: obtaining a target POI to be processed; determining a target recall distance for the target POI based on its target POI type and the recall POI type corresponding to the target POI type; determining a recall location range for the target POI based on the target recall distance and the location information of the target POI; searching for and obtaining the recall POI corresponding to the target POI within the recall location range, and recalling the recall POI.

[0085] Reference Figure 2 , Figure 2This is a flowchart illustrating a point-of-interest (POI) recall method provided in an embodiment of this application. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that shown here. Specifically, the POI recall method may include the following steps 201-204, wherein:

[0086] 201. Obtain the target interest points to be processed.

[0087] To facilitate subsequent understanding, let's first explain points of interest and point of interest retrieval:

[0088] Points of interest (POIs) are geographical entities that have been abstracted into points. For example, POIs can be schools, banks, restaurants, gas stations, hospitals, supermarkets, etc.

[0089] In this application embodiment, a point of interest (POI) can refer to a point of interest on an electronic map. In an electronic map, the information of a POI can include its name, geographical location, and POI type, etc. For example, for the POI "Taihe City Suihanshan Neighborhood Committee Veterans Service Station" on an electronic map, the information it contains can include: the POI name "Taihe City Suihanshan Neighborhood Committee Veterans Service Station," the latitude and longitude of the POI, and the POI type "Community Service Point," etc.

[0090] Among them, electronic maps can refer to maps in map search software that include point-of-interest search functions, and users can download electronic maps through app stores on their smartphones.

[0091] For each point of interest (POI), its POI type is one of several preset types. These preset types can be set manually. When adding POIs to an electronic map, the POI type can be set by manual annotation or automatic identification by computer algorithms.

[0092] The information contained in the points of interest can be stored in the background database of the aforementioned map search software. When it is necessary to obtain the information contained in the points of interest, the electronic device can read it from the background database.

[0093] Points of interest (POIs) retrieval refers to deduplicating duplicate POIs corresponding to the same geographic entity in an electronic map. Since multiple people may repeatedly collect data on the same geographic entity, or the same person may collect data on the same geographic entity multiple times, it is necessary to deduplicate POIs in electronic maps to avoid users getting multiple duplicate POIs when searching for POIs, which can cause confusion. It can also reduce the amount of POI data stored in the background database of map search software.

[0094] In this embodiment of the application, the target point of interest to be processed can refer to a newly added point of interest in an electronic map. In this case, the step "obtain the target point of interest to be processed" includes:

[0095] (1.1) Receive the point of interest entry instruction and obtain the new point of interest corresponding to the point of interest entry instruction.

[0096] Points of interest (POIs) entry commands refer to computer commands used to add points of interest to an electronic map. These commands can be issued by the user. For example, when a user manually collects new POIs, they can issue the command through virtual controls such as "Enter New Point of Interest" in map search software.

[0097] Adding a point of interest (POI) refers to a point of interest added to an electronic map by a user through the POI input command.

[0098] (1.2) Set the newly added interest point as the target interest point to be processed.

[0099] For ease of understanding, unless otherwise stated below, the target point of interest is assumed to be a newly added point of interest, but this should not be construed as a limitation on the embodiments of this application.

[0100] 202. Determine the target recall distance of the target interest point based on the target interest point type and the recall interest point type corresponding to the target interest point type.

[0101] The target interest point type refers to the type of interest point of the target interest point.

[0102] Since the target point of interest (POI) is a newly added POI, its POI type is not stored in the map search software's backend database. Therefore, to obtain the POI type, we can extract the subject words from the POI and use the POI type corresponding to the subject words as the POI type. For example, we can extract the subject words from the POI using Named Entity Recognition (NER) technology. For instance, we can use the open-source 4-bit Sequence Labeling (BMES) model to extract the subject words. Then, based on the subject words in the POI, we can query the pre-defined first subject word library to obtain the corresponding POI type and use that POI type as the POI type.

[0103] Among them, the BMES model is a method for extracting main words in a text by identifying the position of words in the text.

[0104] In this embodiment of the application, the subject term can refer to a word used to characterize a geographic entity. For example, for the point of interest "Taihe City Suihanshan Neighborhood Committee Veterans Service Station", the subject term can be "veterans service station".

[0105] The first subject lexicon stores multiple preset first subject words and their corresponding point-of-interest (POI) types. The first subject words and their corresponding POI types in the lexicon can be manually associated. For example, existing POIs in an electronic map can be pre-acquired, sample subject words can be extracted from these POIs, and the POI types can be obtained. The proportion of POI types corresponding to different sample subject words can be statistically analyzed. POI types with a proportion greater than a preset proportion threshold are taken as the POI types corresponding to the sample subject words. The first subject lexicon is then constructed based on the obtained correspondence. The preset proportion threshold can be set according to actual scenario requirements; for example, it can be set to 90%.

[0106] For example, the first subject term library can be the first backend database of the map search software. When performing step 202, the electronic device can obtain the subject terms in the target point of interest, and then read the first backend database of the map search software to obtain the target point of interest type.

[0107] Since the interest point type differs depending on the position of the subject word within the interest point, a second subject lexicon can be established in some embodiments. This second subject lexicon includes multiple preset second subject words, different word positions corresponding to each second subject word, and different interest point types corresponding to those different word positions. Specifically, the preset second subject lexicon stores multiple preset second subject words, their corresponding different word positions, and their different interest point types corresponding to those different word positions. The second subject words, their word positions, and the different interest point types corresponding to them in the second subject lexicon can be obtained through manual association. The construction method of the second subject lexicon can refer to the construction method of the first subject lexicon, and will not be elaborated further. It is understood that the first and second subject words should be different.

[0108] Among them, word position refers to the location of the main word in the text corresponding to the point of interest.

[0109] For example, the second subject term library can be the second backend database of the map search software. When performing step 202, the electronic device can obtain the subject words and the word positions corresponding to the subject words in the target point of interest, and then read the second backend database of the map search software to obtain the target point of interest type.

[0110] At this point, before the step "determine the target recall distance of the target interest point based on the target interest point type and the recall interest point type corresponding to the target interest point type", the method further includes:

[0111] (2.1) Compare the subject words in the target interest point with the subject words in the preset first database.

[0112] The first database may refer to the aforementioned first main vocabulary database.

[0113] (2.21) If the subject word in the target interest point exists in the first database, then the interest point type corresponding to the subject word in the first database is set as the target interest point type of the target interest point.

[0114] (2.22) If the subject word in the target interest point does not exist in the first database, the interest point type corresponding to the subject word is obtained from the preset second database according to the subject word and the word position of the subject word, and the interest point type is set as the target interest point type of the target interest point.

[0115] The second database can refer to the aforementioned second main vocabulary database.

[0116] In this embodiment of the application, the word position may include one of the following: the beginning of the word, the middle of the word, and the end of the word. The electronic device can use the aforementioned BMES model to extract the main words from the target interest point and simultaneously obtain the word position of the main words.

[0117] The type of recalled interest corresponding to the target point of interest type can refer to the potential type of recalled interest corresponding to the newly added point of interest when the point of interest type is the target point of interest type. Alternatively, since the newly added point of interest and the recalled point of interest correspond to the same geographic entity object, the type of recalled interest corresponding to the target point of interest type can also be understood as the type of interest of historical recalled points of interest corresponding to the geographic entity object corresponding to the target point of interest type, where historical recalled points of interest can refer to points of interest that have been recalled before.

[0118] For example, the type of recalled interest points corresponding to the target interest point type can be statistically obtained through the historical recall records of interest points. For instance, if the target interest point type is "community service point", the recall database used to store historical recall records in the background of the electronic map can be read to obtain the historical recalled interest points corresponding to the historically added interest points when recalling interest points in the past, as well as the interest point types of the historically added interest points and the historically recalled interest points. Then, the historically added interest points of the interest point type "community service point" can be extracted, as well as the target historical recalled interest points corresponding to these historically added interest points. The interest point type of the target historical recalled interest points is taken as the type of recalled interest point corresponding to the target interest point type "community service point".

[0119] The recall database records the historical recall points of interest when adding historical new points of interest, the interest point information of each historical new point of interest, and the interest point information of each historical recall point of interest.

[0120] It should be noted that the number of recall point of interest types can be one or more. The recall point of interest types corresponding to the target point of interest type may or may not include the target point of interest type. For example, when multiple people repeatedly collect data on the geographic entity "XXX4S store", differences in manual settings may lead to different point of interest types for these points of interest. The point of interest types of the multiple points of interest collected repeatedly may include "car repair shop", "car service station", "store", etc. Therefore, in the historical recall process, if the first point of interest with the point of interest type "car repair shop" is a historically added point of interest, and the second point of interest with the point of interest type "car service station" and the third point of interest with the point of interest type "store" are historically recalled points of interest, then the recall point of interest types corresponding to the point of interest type "car repair shop" can include "car service station" and "store". In this case, if the target point of interest type obtained by the electronic device is "car repair shop" when step 202 is executed, then the number of recall point of interest types corresponding to the target point of interest type is 2. If, during the aforementioned historical recall process, there are still historical recall points of interest with the type "auto repair shop," then the recall point of interest types corresponding to the point of interest type "auto repair shop" can include "auto repair shop," "auto service station," and "store." In this case, if the target point of interest type obtained by the electronic device when executing step 202 is "auto repair shop," then the number of recall point of interest types corresponding to the target point of interest type is 3.

[0121] The target recall distance (RDD) refers to the construction parameter used when building the search range for the recall points corresponding to the target interest point. It characterizes the size of the search range used when searching for the recall points corresponding to the target interest point. A larger RDD indicates a larger search range, while a smaller RDD indicates a smaller search range.

[0122] It should be noted that the target recall distance can refer to either the virtual distance on an electronic map or the actual distance. For ease of understanding, unless otherwise specified below, the target recall distance is assumed to be the actual distance, but this should not be construed as a limitation on the embodiments of this application.

[0123] In some embodiments, a preset parameter table can be queried based on the target point of interest type and the recalled point of interest type to obtain the target recall distance of the target point of interest. The preset parameter table can be manually set and stored in the background database of the map search software. The parameter table includes the correspondence between the point of interest type of the newly added point of interest, the corresponding recalled point of interest type, and the recall distance of the newly added point of interest. During step 202, the electronic device reads the parameter table from the background database of the map search software and queries the target recall distance of the target point of interest based on the target point of interest type and the recalled point of interest type.

[0124] Understandably, the number of recall point of interest types should be the same as the number of target recall distances. For example, for the target point of interest type "auto repair shop," if its corresponding recall point of interest types are "auto service station" and "store," then 2 target recall distances can be obtained. However, when the target point of interest type "auto repair shop" corresponds to the recall point of interest types "auto repair shop," "auto service station," and "store," then 3 target recall distances can be obtained.

[0125] In this embodiment, the purpose of obtaining the recall point of interest type corresponding to the target point of interest type and determining the target recall distance through the target point of interest type and the recall point of interest type is to narrow the search range when searching for recall points of interest. If the similarity between newly added points of interest and existing points of interest already included in the electronic map is determined solely by the similarity between points of interest, and recall points of interest are selected from existing points of interest based on the determination result, then the similarity between the newly added point of interest and each existing point of interest needs to be calculated. When the electronic map contains a large number of existing points of interest and the frequency of adding points of interest is high, the required computational load is extremely large. Here, the similarity between points of interest can be understood as the similarity between the names of points of interest. For example, open-source language processing models such as word2vec can be used to convert the names of the points of interest to be calculated into vectors, and then the similarity between the vectors corresponding to the points of interest can be calculated.

[0126] The following explains the logic for determining the target recall distance based on the target point of interest and the type of recalled point of interest: Since different geographic entities occupy different geographic areas, and data collectors typically use their own location as the location of the collected points of interest, the distance range of duplicate points of interest obtained when manually collecting points of interest corresponding to different geographic entities will vary. Consequently, the search range for recalling points of interest will also differ. By combining the type of target point of interest and the corresponding type of recalled point of interest, we can characterize the type of geographic entity corresponding to the target point of interest. Then, based on the type of geographic entity, we can determine the target recall distance of the target point of interest to construct different search ranges.

[0127] For example, using the example above, suppose the target point of interest type is "car repair shop", and the corresponding recall point of interest type includes "car service station". From the target point of interest type and the recall point of interest type, we know that the geographic entity object corresponding to the target point of interest may be a 4S store. For the geographic entity object 4S store, since the geographic entity object occupies a small geographic area, even if the collectors stand at different positions such as the front door and back door of the geographic entity object when manually collecting the points of interest of the geographic entity object, the distance between the collected points of interest will not be too large. Therefore, the distance between the duplicate points of interest of the geographic entity object is small, the target recall distance is small, and the constructed search range is small.

[0128] For comparison, let's make another assumption: Suppose the target point of interest type is "residential," and the corresponding recall point of interest types include "villa area." Based on the target point of interest type and the recall point of interest type, we know that the geographic entity object corresponding to the target point of interest may be a residential community. For a residential community, since the geographic entity object occupies a large geographic area, when manually collecting points of interest for this geographic entity object, if the collector stands at different positions such as the front door, back door, and center of the geographic entity object, the distance between the collected points of interest will be large. Therefore, the distance between repeated points of interest for this geographic entity object will be large, resulting in a larger target recall distance and a larger search range.

[0129] As can be seen, the embodiments of this application can obtain all possible types of geographic entity objects corresponding to the target point of interest by using the target point of interest type and the recall point of interest type. Then, for each type of geographic entity object, the corresponding target recall distance is output to determine the corresponding search range, which can greatly reduce the amount of search calculation during the search.

[0130] 203. Based on the target recall distance and the location information of the target point of interest, determine the recall location range of the target point of interest.

[0131] For an explanation of the location information of points of interest, please refer to step 201; details will not be elaborated further.

[0132] The recall location range refers to the range of points of interest that can be searched and recalled. When the target recall distance is the actual distance, the recall location range refers to the actual geographical range.

[0133] The recall location range for a target point of interest can be constructed in several ways. The following explanation focuses on the case where there are multiple target recall distances:

[0134] In some embodiments, the electronic device can select the maximum recall distance from the target recall distance, use the maximum recall distance as the radius, and use the location of the target point of interest as the center to construct a circular recall location range, so as to maximize the search range and avoid missing recall points of interest.

[0135] In other embodiments, the electronic device can select the maximum recall distance from the target recall distance, use twice the maximum recall distance as the side length, and use the location of the target point of interest as the center to construct a square recall location range, so as to maximize the search range and avoid missing the recall point of interest.

[0136] This application does not impose any restrictions on the method of constructing the recall location range.

[0137] In some embodiments, it can also be determined whether the geographic entity object corresponding to the target point of interest is a chain brand. When the geographic entity object corresponding to the target point of interest is a chain brand, different recall location range construction strategies are adopted. In this case, the step "determine the recall location range of the target point of interest based on the target recall distance and the location information of the target point of interest" includes:

[0138] (A) If the target interest point includes a target subject word, then the recall location range of the target interest point is determined based on the minimum target recall distance and the location information of the target interest point, wherein the target subject word is a subject word in a preset chain brand database.

[0139] A chain brand database refers to a database that stores the main keywords corresponding to chain brands.

[0140] The chain brand database stores the main keywords corresponding to all chain brands. When retrieving data from this database, one can first use a search engine to obtain all chain brands from the internet, and then extract the corresponding main keywords from these brands. For example, one can use a search engine to retrieve all chain brand names from the internet, then extract the corresponding main keywords, and store these main keywords on a cloud server, which can serve as the chain brand database.

[0141] If the target point of interest contains the main keywords from the chain brand database, then the geographic entity corresponding to the target point of interest can be determined to be a chain brand. Since the point of interest names corresponding to geographic entities of the same chain brand are usually quite similar, if the search range for recalling points of interest is too large, it may mistakenly recall non-recallable points of interest. To avoid these problems, the search range for recalling points of interest can be narrowed as much as possible, that is, the recall location range can be constructed based on the minimum target recall distance.

[0142] (B) If the target interest point does not include the target subject word, then the recall location range of the target interest point is determined based on the maximum target recall distance and the location information of the target interest point.

[0143] If the geographic entity corresponding to the target point of interest is not a chain brand, then in order to maximize the search scope and avoid missing the recall points of interest, the recall location range can be constructed based on the maximum target recall distance.

[0144] 204. Search within the recall location range to obtain the recall point of interest corresponding to the target point of interest, and recall the recall point of interest.

[0145] During the search, the similarity between the target point of interest and all preset points of interest within the recall location range can be compared. Preset points of interest with a similarity greater than a preset similarity threshold are used as the recall points of interest corresponding to the target point of interest, and the recalled points of interest are recalled. For example, the information of the recalled points of interest can be deleted from the background database of the map search software.

[0146] The method for calculating similarity can be found in the explanation above, and will not be elaborated further here.

[0147] Among them, preset points of interest refer to points of interest that already exist in the electronic map.

[0148] In summary, the point of interest recall method provided in this application includes: acquiring a target point of interest to be processed; determining a target recall distance for the target point of interest based on the target point of interest type and the recall point of interest type corresponding to the target point of interest type; determining a recall location range for the target point of interest based on the target recall distance and the location information of the target point of interest; searching for and retrieving the recall point of interest corresponding to the target point of interest within the recall location range; and recalling the recall point of interest.

[0149] As can be seen, the point of interest recall method provided in this application obtains the target recall distance by using the target point of interest type and the recall point of interest type, and then constructs the recall location range based on the target recall distance. Therefore, while limiting the search range when searching for recall points of interest, it improves the accuracy of the recall location range. This reduces the number of points of interest that need to be searched when searching for recall points of interest, reduces the computational load of point of interest recall, improves the efficiency of point of interest recall, and ensures the accuracy of obtaining recall points of interest, avoiding missed selections and incorrect selections.

[0150] In some embodiments, the electronic device can also obtain the target recall distance through a trained machine learning model to improve accuracy. (Reference) Figure 3 At this point, the step "determine the target recall distance of the target interest point based on the target interest point type and the recall interest point type corresponding to the target interest point type" includes:

[0151] 301. Obtain the target interest point type, the recall interest point type corresponding to the target interest point type, the first word vector corresponding to the target interest point type, and the second word vector corresponding to the recall interest point type.

[0152] Among them, electronic devices can use open-source language processing models such as word2vec to convert text of the target interest type into the first word vector and text of the recall interest type into the second word vector.

[0153] The methods for obtaining target interest type and recall interest type can be found in the above text, and will not be elaborated here.

[0154] 302. Based on the target interest point type and the corresponding recall interest point type, select a recall distance prediction model from the preset candidate models.

[0155] The preset candidate model is a pre-trained machine learning model used to predict the recall distance. In some embodiments of this application, a logistic regression model can be used as the initial candidate model, and the initial candidate model can be trained using sample data to obtain the preset candidate model.

[0156] To improve accuracy, a preset candidate model can be used for each target point of interest type and recall point of interest type. For example, the first candidate model can be used for the target point of interest type "car repair shop" and the recall point of interest type "car service station", while the second candidate model can be used for the target point of interest type "car repair shop" and the recall point of interest type "shop".

[0157] The sample data may include the sample interest point type, the corresponding recall interest point type, and the actual recall distance, which can be obtained by manual annotation.

[0158] During training, a first sample vector of the sample interest point type and a second sample vector of the recall interest point type corresponding to the sample interest point type can be obtained. Then, the first sample vector and the second sample vector are fused and input into the initial candidate model. The predicted recall distance is obtained by using the objective function of the initial candidate model. Based on the predicted recall distance and the actual recall distance, the initial candidate model is trained to obtain the preset candidate model.

[0159] The objective function of the initial candidate model can be a normal distribution function, as shown in equation (1):

[0160]

[0161] Where f(x) is the recall distance, σ and μ are the parameters to be trained, which can be understood as the standard deviation and expected value of the normal distribution, respectively, and x is the fusion vector of the first sample vector and the second sample vector.

[0162] The fusion methods can include splicing, averaging, and other methods.

[0163] The loss function for the initial candidate model can be the cross-entropy loss function.

[0164] It should be noted that for each initial candidate model, the types of interest points included in each sample data used during training should be the same, and the types of recall interest points should also be the same. Specifically, for the initial first candidate model, the target interest point type should be "car repair shop," and the recall interest point type should be "car service station" in each sample data used during training; similarly, for the initial second candidate model, the target interest point type should be "car repair shop," and the recall interest point type should be "shop." Through training, the values ​​of σ and μ can be obtained for each target interest point type and recall interest point type. In other words, the objective function in each pre-defined candidate model can be used to characterize the normal distribution of the recall distance under the corresponding target interest point type and recall interest point type. It is understandable that the greater the distance between two points of interest, the lower the probability of the two points of interest being repeated, and the lower the probability of recall. Therefore, by using the trained formula (1), the maximum f(x), that is, the recall distance with the lowest probability of recall, can be used as the range when searching for and recalling points of interest. Thus, the integrity of searching for and recalling points of interest can be guaranteed, and no omissions will occur.

[0165] Furthermore, if, as described in step 202, the target recall distance is obtained by "querying a preset parameter table", then the parameter table can be constructed by "training an initial candidate model, obtaining a preset candidate model, and predicting the data in the parameter table using the preset candidate model", and the parameter table can be stored in the background database of the map search software. When executing step 202, the electronic device can read and query the parameter table.

[0166] 303. After fusing the first word vector and the second word vector, input them into the recall distance prediction model to predict the target recall distance of the target interest point.

[0167] As mentioned above, since the recall distance prediction model can characterize the normal distribution of the recall distance under the corresponding target interest type and recall interest type, that is, under the corresponding target interest type and recall interest type, the recall distance best matches the actual distribution. Therefore, if the first word vector and the second word vector are fused together, the target recall distance can be obtained.

[0168] In addition to constructing only one recall location range for point of interest search as described in step 203, two different shapes of recall location ranges can be constructed simultaneously. Combining the advantages of different shapes reduces computational load while still obtaining the recalled points of interest. (Reference) Figure 4At this time, the recall location range includes a first location range and a second location range. The first location range is rectangular, and the area of ​​the second location range is smaller than the area of ​​the first location range. The step "searching within the recall location range to obtain the recall point of interest corresponding to the target point of interest, and recalling the recall point of interest" includes:

[0169] 401. Obtain the first candidate point of interest whose location is within the range of the first location.

[0170] In this embodiment of the application, the first location range may refer to the square recall location range in step 203.

[0171] In some embodiments, the first candidate point of interest may refer to all preset points of interest within the first location range. The description of the preset points of interest can be found in step 204, and will not be elaborated further.

[0172] In other embodiments, the electronic device can also filter and obtain first candidate points of interest based on the recording time. In this case, the step "obtaining first candidate points of interest whose locations are within the first location range" includes:

[0173] (3.1) Take the initial points of interest located within the recall location range as the points of interest to be screened, and obtain the recording time of the points of interest to be screened.

[0174] At this point, the points of interest to be filtered refer to all preset points of interest within the first position range.

[0175] The entry duration refers to the time elapsed since the point of interest (POI) was added to the electronic map up to the current time of step (3.1). The entry duration of a POI can be calculated based on its entry time and the current time. The entry time is stored in the background database of the map search software along with the POI information at the time of entry. When step (3.1) is executed, the electronic device can retrieve this information from the background database.

[0176] (3.2) The interest points to be screened are screened according to the recording duration to obtain the first candidate interest points with a recording duration greater than or equal to the preset time threshold.

[0177] For example, the electronic device can exclude points of interest to be screened whose recording duration is less than a preset time threshold, and select points of interest to be screened whose recording duration is greater than or equal to the preset time threshold as the first candidate points of interest.

[0178] The reason for the above filtering is that usually not many people collect points of interest for the same geographic entity within a short period of time.

[0179] The reason for filtering preset interest points by the first position range is that determining whether a preset interest point is within the first position range is faster than determining whether it is within the position range of other shapes. Therefore, the preset interest points can be coarsely screened to obtain the first candidate interest points.

[0180] The reason why filtering through the first location range is faster is that, since the first location range is a square range, when determining whether a preset point of interest (POI) is within the first location range, it is only necessary to calculate the positional difference between the preset POI and the target POI, and the relationship between this difference and the side length of the first location range (i.e., twice the target recall distance). For example, if the latitude and longitude of the preset POI are (114.0431°, 22.6742°), and the latitude and longitude of the target POI are (114.0418°, 22.6735°), and the first location... When the side length of the range is converted to longitude, it corresponds to a positional difference of 0.03°. When converted to latitude, it corresponds to a positional difference of 0.03°. Since the positional difference between the preset point of interest (POI) and the target POI is (0.0013°, 0.0007°), the positional difference between the preset POI and the target POI is less than half the positional difference corresponding to the side length in both longitude and latitude (i.e., the target recall distance). Therefore, it can be determined that the preset POI is within the first positional range. However, when using positional ranges of other shapes, such as circular ranges, it is necessary to perform a Pythagorean theorem calculation on the positional difference between the preset POI and the target POI, and then compare the calculation result with the radius of the circular range (i.e., the target recall range) before making a judgment. It can be seen that through step 401, the electronic device can first perform a coarse screening of all preset POIs in the electronic map using a method with less computational load to obtain the first candidate POI.

[0181] 402. Compare the similarity between the first candidate interest point and the target interest point with a preset similarity threshold to obtain a second candidate interest point with a similarity greater than the similarity threshold.

[0182] After the initial screening, a finer screening can be performed using a similarity comparison method to obtain the second candidate interest points.

[0183] It should be noted that if a second candidate point of interest with a similarity greater than the similarity threshold cannot be found, it may be due to improper user input of the target point of interest, resulting in typos or omissions in the name, or the geographic entity corresponding to the target point of interest may have been renamed. Therefore, for candidate points of interest that correspond to the same geographic entity as the target point of interest, their similarity to the target point of interest is relatively low. In this case, the first subject word in the first candidate point of interest and the second subject word in the target point of interest can be extracted separately, and the similarity between the first subject word and the second subject word can be used as the similarity between the first candidate point of interest and the target point of interest to avoid the above problems.

[0184] 403. Set the second candidate point of interest whose location is within the second location range as the recall point of interest corresponding to the target point of interest.

[0185] To avoid an excessively large search range for recalling points of interest, which could lead to the inclusion of non-recallable points of interest in the second candidate points of interest, the electronic device can further filter the second candidate points of interest using a smaller second location range. For example, the electronic device can use the circular recall location range from step 203 as the second location range. It is understood that in the square and circular recall location ranges constructed in step 203, the area of ​​the circular recall location range is smaller than that of the square recall location range. Therefore, the method in step 203 can be used to simultaneously construct both a square and a circular recall location range, using the square recall location range as the first location range and the circular recall location range as the second location range, without needing to process the target recall distance according to other construction rules and reconstruct the first and second location ranges. This reduces computational load, improves the efficiency of point of interest recall, and saves storage space used to store the aforementioned construction rules.

[0186] 404. Recall the points of interest mentioned above.

[0187] To better implement the point-of-interest (POI) recall method in the embodiments of this application, this application also provides a POI recall device based on the POI recall method, such as... Figure 5 The diagram shown is a structural schematic of one embodiment of the point-of-interest (POI) recall device in this application. The POI recall device 500 includes:

[0188] Acquisition unit 501 is used to acquire target interest points to be processed;

[0189] The distance determination unit 502 is used to determine the target recall distance of the target interest point based on the target interest point type and the recall interest point type corresponding to the target interest point type.

[0190] The range determination unit 503 is used to determine the recall location range of the target interest point based on the target recall distance and the location information of the target interest point;

[0191] The recall unit 504 is used to search for and retrieve the recall point of interest corresponding to the target point of interest within the recall location range, and to retrieve the recall point of interest.

[0192] In one possible implementation of this application, the distance determining unit 502 is further configured to:

[0193] Obtain the target interest point type, the recall interest point type corresponding to the target interest point type, the first word vector corresponding to the target interest point type, and the second word vector corresponding to the recall interest point type;

[0194] Based on the target interest point type and the corresponding recall interest point type, a recall distance prediction model is selected from the preset candidate models.

[0195] The first word vector and the second word vector are fused and input into the recall distance prediction model to predict the target recall distance of the target interest point.

[0196] In one possible implementation of this application, the range determination unit 503 is further configured to:

[0197] The step of determining the recall location range of the target interest point based on the target recall distance and the location information of the target interest point includes:

[0198] If the target interest point includes a target subject word, then the recall location range of the target interest point is determined based on the minimum target recall distance and the location information of the target interest point, wherein the target subject word is a subject word in a preset chain brand database;

[0199] If the target interest point does not include the target subject word, then the recall location range of the target interest point is determined based on the maximum target recall distance and the location information of the target interest point.

[0200] In one possible implementation of this application, the distance determining unit 502 is further configured to:

[0201] The subject words in the target interest points are compared with the subject words in a preset first database, wherein the first database contains the interest point types corresponding to the preset first subject words;

[0202] If the subject word in the target interest point exists in the first database, then the interest point type corresponding to the subject word in the first database is set as the target interest point type of the target interest point;

[0203] If the subject word in the target interest point does not exist in the first database, then based on the subject word and its word position, the interest point type corresponding to the subject word is obtained from a preset second database, and the interest point type is set as the target interest point type of the target interest point. The word position includes one of the beginning, middle, and end of the word, and the second database contains preset interest point types corresponding to the second subject word in different word positions.

[0204] In one possible implementation of this application, the recall unit 504 is further configured to:

[0205] Obtain the first candidate point of interest whose location is within the range of the first location;

[0206] The similarity between the first candidate interest point and the target interest point is compared with a preset similarity threshold to obtain a second candidate interest point with a similarity greater than the similarity threshold.

[0207] The second candidate point of interest whose location is within the second location range is set as the recall point of interest corresponding to the target point of interest;

[0208] The recall refers to the points of interest mentioned above.

[0209] In one possible implementation of this application, the recall unit 504 is further configured to:

[0210] The initial points of interest located within the recall location range are used as points of interest to be filtered, and the entry time of the points of interest to be filtered is obtained.

[0211] The points of interest to be screened are filtered based on the entry time to obtain the first candidate points of interest.

[0212] The initial points of interest located within the recall location range are used as points of interest to be filtered, and the recording time of the points of interest to be filtered is obtained.

[0213] The points of interest to be screened are filtered based on the recording duration to obtain the first candidate points of interest with a recording duration greater than or equal to a preset time threshold.

[0214] In one possible implementation of this application, the acquiring unit 501 is further configured to:

[0215] Receive an input instruction for points of interest and obtain the newly added points of interest corresponding to the input instruction;

[0216] The newly added point of interest is set as the target point of interest to be processed.

[0217] In practice, each of the above units can be implemented as an independent entity or can be arbitrarily combined to be implemented as the same or several entities. For the specific implementation of each of the above units, please refer to the previous method embodiments, which will not be repeated here.

[0218] Since this point-of-interest recall device can execute the steps in the point-of-interest recall method in any embodiment, it can achieve the beneficial effects that the point-of-interest recall method in any embodiment of this application can achieve, as detailed in the preceding description, and will not be repeated here.

[0219] Furthermore, in order to better implement the point of interest recall method in the embodiments of this application, the point of interest recall method is...

[0220] Based on this, embodiments of this application also provide an electronic device, see below. Figure 6 , Figure 6 This illustration shows a structural diagram of an electronic device according to an embodiment of this application. Specifically, the electronic device provided in this embodiment includes a processor 601. The processor 601 is used to execute a computer program stored in a memory 602 to implement the steps of the interest point recall method in any embodiment; or, the processor 601 is used to execute a computer program stored in a memory 602 to implement, for example... Figure 5 The functions of each unit in the corresponding embodiment.

[0221] For example, a computer program may be divided into one or more modules / units, one or more of which are stored in memory 602 and executed by processor 601 to complete the embodiments of this application. One or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in a computer device.

[0222] The electronic device may include, but is not limited to, processor 601 and memory 602. Those skilled in the art will understand that the illustrations are merely examples of an electronic device and do not constitute a limitation on the device. It may include more or fewer components than illustrated, or combine certain components, or use different components.

[0223] Processor 601 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the electronic device, connecting various parts of the electronic device through various interfaces and lines.

[0224] The memory 602 can be used to store computer programs and / or modules. The processor 601 implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory 602 and by calling data stored in the memory 602. The memory 602 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device (such as audio data, video data, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0225] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the point of interest recall device, electronic device and its corresponding unit described above can be referred to the description of the point of interest recall method in any embodiment, and will not be repeated here.

[0226] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be accomplished by instructions, or by instructions controlling related hardware. These instructions can be stored in a storage medium and loaded and executed by a processor.

[0227] Therefore, embodiments of this application provide a storage medium storing a computer program. When the computer program is executed by a processor, it performs the steps of the point of interest recall method in any embodiment of this application. For specific operations, please refer to the description of the point of interest recall method in any embodiment, which will not be repeated here.

[0228] The storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0229] Since the instructions stored in the storage medium can execute the steps in the point of interest recall method in any embodiment of this application, the beneficial effects that the point of interest recall method in any embodiment of this application can achieve can be realized, as detailed in the preceding description, and will not be repeated here.

[0230] The above provides a detailed description of a point-of-interest recall method, apparatus, storage medium, and electronic device provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for recalling points of interest, characterized in that, include: Obtain the target interest points to be processed; Based on the target interest point type and the corresponding recall interest point type, determine the target recall distance of the target interest point. Based on the target recall distance and the location information of the target point of interest, the recall location range of the target point of interest is determined; there are at least two target recall distances, and the method for determining the recall location range includes: If the target interest point includes a target subject word, then the recall location range of the target interest point is determined based on the minimum target recall distance and the location information of the target interest point, wherein the target subject word is a subject word in a preset chain brand database; If the target interest point does not include the target subject word, then the recall location range of the target interest point is determined based on the maximum target recall distance and the location information of the target interest point. Within the recall location range, the recall point of interest corresponding to the target point of interest is obtained, and the recall point of interest is recalled.

2. The point-of-interest recall method according to claim 1, characterized in that, The step of determining the target recall distance of the target interest point based on the target interest point type and the corresponding recall interest point type includes: Obtain the target interest point type, the recall interest point type corresponding to the target interest point type, the first word vector corresponding to the target interest point type, and the second word vector corresponding to the recall interest point type; Based on the target interest point type and the corresponding recall interest point type, a recall distance prediction model is selected from the preset candidate models. The first word vector and the second word vector are fused and input into the recall distance prediction model to predict the target recall distance of the target interest point.

3. The point-of-interest recall method according to claim 1, characterized in that, Before determining the target recall distance of the target interest point based on its target interest point type and the corresponding recall interest point type, the method further includes: The subject words in the target interest points are compared with the subject words in a preset first database, wherein the first database contains the interest point types corresponding to the preset first subject words; If the subject word in the target interest point exists in the first database, then the interest point type corresponding to the subject word in the first database is set as the target interest point type of the target interest point; If the subject word in the target interest point does not exist in the first database, then based on the subject word and its word position, the interest point type corresponding to the subject word is obtained from a preset second database, and the interest point type is set as the target interest point type of the target interest point. The word position includes one of the beginning, middle, and end of the word, and the second database contains preset interest point types corresponding to the second subject word in different word positions.

4. The point-of-interest recall method according to claim 1, characterized in that, The recall location range includes a first location range and a second location range. The first location range is rectangular, and the area of ​​the second location range is smaller than the area of ​​the first location range. The step of searching for and retrieving the recall point of interest corresponding to the target point of interest within the recall location range, and recalling the recall point of interest, includes: Obtain the first candidate point of interest whose location is within the range of the first location; The similarity between the first candidate interest point and the target interest point is compared with a preset similarity threshold to obtain a second candidate interest point with a similarity greater than the similarity threshold. The second candidate point of interest whose location is within the second location range is set as the recall point of interest corresponding to the target point of interest; The recall refers to the points of interest mentioned above.

5. The point-of-interest recall method according to claim 4, characterized in that, The acquisition of the first candidate interest point located within the first location range includes: The initial points of interest located within the recall location range are used as points of interest to be filtered, and the recording time of the points of interest to be filtered is obtained. The points of interest to be screened are filtered based on the recording duration to obtain the first candidate points of interest with a recording duration greater than or equal to a preset time threshold.

6. The point-of-interest recall method according to any one of claims 1-5, characterized in that, The acquisition of the target interest points to be processed includes: Receive an input instruction for points of interest and obtain the newly added points of interest corresponding to the input instruction; The newly added point of interest is set as the target point of interest to be processed.

7. A point-of-interest recall device, characterized in that, include: The acquisition unit is used to acquire the target interest points to be processed. The distance determination unit is used to determine the target recall distance of the target interest point based on the target interest point type and the recall interest point type corresponding to the target interest point type. A range determination unit is used to determine the recall location range of the target point of interest based on the target recall distance and the location information of the target point of interest; the target recall distance is at least two, and the method for determining the recall location range includes: If the target interest point includes a target subject word, then the recall location range of the target interest point is determined based on the minimum target recall distance and the location information of the target interest point, wherein the target subject word is a subject word in a preset chain brand database; If the target interest point does not include the target subject word, then the recall location range of the target interest point is determined based on the maximum target recall distance and the location information of the target interest point. The recall unit is used to search for and retrieve the recall point of interest corresponding to the target point of interest within the recall location range.

8. An electronic device, characterized in that, The electronic device includes a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the point-of-interest recall method as described in any one of claims 1 to 6.

9. A storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the point-of-interest recall method according to any one of claims 1 to 6.

Images