Point of interest recommendation method, apparatus, device, and storage medium

Abstract

The application relates to the technical field of Internet, and discloses a point of interest recommendation method, device, equipment and storage medium, the method comprises the following steps: obtaining a candidate point of interest based on the current position of a target terminal; obtaining user historical behavior data sent by the target terminal, and determining a user preference center word according to the user historical behavior data; selecting a recommended point of interest from the candidate point of interest according to the user preference center word, and pushing the recommended point of interest to the target terminal for display. Unlike the prior art which only uses user search words to recommend points of interest, the application determines a user preference center word according to user historical behavior data of a target terminal, and then selects a recommended point of interest from candidate points of interest obtained based on the current position of the target terminal according to the user preference center word, so that the accuracy of the obtained recommended point of interest is improved, and further, user stickiness and user experience are improved.

Description

Technical Field

[0001] This invention relates to the field of Internet technology, and in particular to a method, apparatus, device, and storage medium for recommending points of interest. Background Technology

[0002] With the rapid development of internet technology, online resources have also grown exponentially. Traditional general-purpose search engines rely on user-inputted search terms. However, in practice, even with the same search terms, different users may have different search objectives and desired results. Furthermore, traditional search recommendation methods based on search logs cannot quickly and effectively process massive amounts of log information, nor can they capture user interests. Therefore, to increase user engagement, Point of Interest (POI) recommendation services have emerged. These services aim to recommend points of interest that users may be interested in, providing supplementary information during the user's search process. However, existing POI recommendation services have low accuracy and still struggle to meet users' personalized service needs. Therefore, improving the accuracy of POI recommendations to enhance user engagement and user experience has become an urgent problem to solve.

[0003] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main objective of this invention is to provide a method, apparatus, device, and storage medium for recommending points of interest, aiming to solve the technical problem of how to improve the accuracy of point of interest recommendations in order to enhance user stickiness and user experience.

[0005] To achieve the above objectives, the present invention provides a point-of-interest recommendation method, the method comprising the following steps:

[0006] Obtain candidate points of interest based on the current location of the target terminal;

[0007] Obtain user historical behavior data sent by the target terminal, and determine user preference keywords based on the user historical behavior data of the target terminal;

[0008] Based on the user preference keywords, select the points of interest to be recommended from the candidate points of interest, and push the points of interest to be recommended to the target terminal for display.

[0009] Optionally, the step of obtaining candidate points of interest based on the current location of the target terminal specifically includes:

[0010] The region where the target terminal is located is obtained, and the region where the terminal is located is divided into corresponding regional grids through map projection;

[0011] Query the target area grid corresponding to the current location of the target terminal in the area grid, and obtain the candidate points of interest corresponding to the target area grid.

[0012] Optionally, the step of obtaining the area where the target terminal is located and dividing the area into corresponding regional grids through map projection specifically includes:

[0013] Obtain the target terminal's location area and the current scene information of the location area;

[0014] The corresponding region segmentation rule is matched based on the current scene information, and the region where the terminal is located is segmented into a corresponding region grid through map projection according to the region segmentation rule.

[0015] Optionally, the step of obtaining user historical behavior data sent by the target terminal and obtaining user preference keywords based on the user historical behavior data of the target terminal specifically includes:

[0016] Obtain the user's historical behavior data sent by the target terminal, and convert the user's historical behavior data into corresponding word vectors;

[0017] Calculate the similarity between the word vectors, and determine the user preference center word based on the similarity.

[0018] Optionally, the step of obtaining the user's historical behavior data sent by the target terminal and converting the user's historical behavior data into corresponding word vectors specifically includes:

[0019] The system acquires user historical behavior data sent by the target terminal and performs word segmentation on the user historical behavior data to obtain a vocabulary set.

[0020] The central word is determined based on the vocabulary set, and the central word is mapped to a vector space to obtain the corresponding word vector.

[0021] Optionally, the step of calculating the similarity between the word vectors and determining the user preference center word based on the similarity specifically includes:

[0022] Calculate the similarity between the word vectors, and sort the similarity to obtain the similarity ranking result;

[0023] User-preferred central words are selected from the central words based on the similarity ranking results.

[0024] Optionally, before the step of obtaining the candidate points of interest based on the current location of the target terminal, the method further includes:

[0025] Obtain the globally unique identifier of the current user of the target terminal, and determine whether the current user is an unregistered user based on the globally unique identifier;

[0026] If the current user is not the unregistered user, the step of obtaining the candidate points of interest based on the current location of the target terminal is executed.

[0027] Optionally, after the step of obtaining the globally unique identifier of the current user of the target terminal and determining whether the current user is an unregistered user based on the globally unique identifier, the method further includes:

[0028] When the current user is the unregistered user, obtain the region where the target terminal is located;

[0029] The system queries the list of recommended regional points of interest (POIs) corresponding to the region where the terminal is located, and pushes the POIs recommendation list to the target terminal for display.

[0030] Optionally, before the step of obtaining the user's historical behavior data sent by the target terminal and determining the user preference center words based on the user's historical behavior data of the target terminal, the method further includes:

[0031] Determine whether the target terminal has received the current user's historical behavior data;

[0032] If the user's historical behavior data is not received from the target terminal, the region where the target terminal is located is obtained.

[0033] The system queries the list of recommended regional points of interest (POIs) corresponding to the region where the terminal is located, and pushes the POIs recommendation list to the target terminal for display.

[0034] Optionally, before the step of querying the regional point of interest recommendation list corresponding to the region where the terminal is located and pushing the regional point of interest recommendation list to the target terminal for display, the method further includes:

[0035] Obtain the regional candidate points of interest in different regions and the ranking factors corresponding to the regional candidate points of interest;

[0036] The sorting factor is normalized to obtain the normalized sorting factor.

[0037] The candidate interest points in the region are sorted according to the normalized sorting factor to generate a recommended list of regional interest points for different regions.

[0038] Optionally, before the step of obtaining the regional candidate points of interest in different regions and the ranking factors corresponding to the regional candidate points of interest, the method further includes:

[0039] Obtain historical behavior data for different regions, and extract regional behavior items from the historical behavior data.

[0040] The number of clicks on the regional behavior items is counted, and the click counts are sorted to generate a regional behavior ranking result;

[0041] Based on the ranking results of the regional behaviors, select candidate interest points for different regions from the regional behavior items.

[0042] Optionally, the step of selecting points of interest to be recommended from the candidate points of interest based on the user preference keywords, and pushing the points of interest to be recommended to the target terminal for display, specifically includes:

[0043] Based on the user preference keywords, select the interest points to be recommended from the candidate interest points, and generate an interest point recommendation list based on the interest points to be recommended;

[0044] The list of points of interest recommendations is pushed to the target terminal for adaptive display.

[0045] Furthermore, to achieve the above objectives, the present invention also proposes an interest point recommendation device, the interest point recommendation device comprising:

[0046] The candidate point of interest acquisition module is used to acquire candidate points of interest based on the current location of the target terminal;

[0047] The user preference acquisition module is used to acquire user historical behavior data sent by the target terminal, and determine user preference core words based on the user historical behavior data of the target terminal.

[0048] The point of interest recommendation module is used to select points of interest to be recommended from the candidate points of interest based on the user's preference keywords, and push the points of interest to be recommended to the target terminal for display.

[0049] Optionally, the candidate point of interest acquisition module is further configured to acquire the area where the target terminal is located, and to divide the area where the terminal is located into corresponding area grids through map projection.

[0050] The candidate point of interest acquisition module is further configured to query the target area grid corresponding to the current location of the target terminal in the area grid, and acquire the candidate point of interest corresponding to the target area grid.

[0051] Optionally, the candidate point of interest acquisition module is further configured to acquire the area where the target terminal is located and the current scene information of the area where the terminal is located;

[0052] The candidate point of interest acquisition module is further configured to match the corresponding region segmentation rule according to the current scene information, and to segment the region where the terminal is located into a corresponding region grid through map projection according to the region segmentation rule.

[0053] Optionally, the user preference acquisition module is further configured to acquire user historical behavior data sent by the target terminal and convert the user historical behavior data into corresponding word vectors;

[0054] The user preference acquisition module is also used to calculate the similarity between the word vectors and determine the user preference center word based on the similarity.

[0055] Optionally, the user preference acquisition module is further configured to acquire user historical behavior data sent by the target terminal, and perform word segmentation processing on the user historical behavior data to obtain a vocabulary set;

[0056] The user preference acquisition module is further configured to determine the central word based on the vocabulary set and map the central word to a vector space to obtain the corresponding word vector.

[0057] Optionally, the user preference acquisition module is further configured to calculate the similarity between the word vectors and sort the similarity to obtain a similarity ranking result;

[0058] The user preference acquisition module is further configured to select user preference center words from the center words based on the similarity ranking results.

[0059] Furthermore, to achieve the above objectives, the present invention also proposes an interest point recommendation device, the device comprising: a memory, a processor, and an interest point recommendation program stored in the memory and executable on the processor, the interest point recommendation program being configured to implement the steps of the interest point recommendation method as described above.

[0060] Furthermore, to achieve the above objectives, the present invention also proposes a storage medium storing an interest point recommendation program, which, when executed by a processor, implements the steps of the interest point recommendation method as described above.

[0061] In this invention, candidate points of interest (POIs) are obtained based on the current location of the target terminal. User historical behavior data sent by the target terminal is also acquired, and user preference keywords are determined based on this data. POIs to be recommended are then selected from the candidate POIs based on these user preference keywords and pushed to the target terminal for display. Unlike existing technologies that rely solely on user search terms for POI recommendations, resulting in low recommendation relevance, this invention improves the accuracy of the acquired POIs by acquiring user historical behavior data sent by the target terminal, determining user preference keywords based on this data, and then selecting POIs to be recommended from the candidate POIs obtained based on the target terminal's current location. This further enhances user engagement and user experience. Attached Figure Description

[0062] Figure 1 This is a schematic diagram of the structure of the point-of-interest recommendation device in the hardware operating environment involved in the embodiments of the present invention;

[0063] Figure 2 This is a flowchart illustrating the first embodiment of the point-of-interest recommendation method of the present invention;

[0064] Figure 3 This is a flowchart illustrating the second embodiment of the point of interest recommendation method of the present invention;

[0065] Figure 4 This is a flowchart illustrating the third embodiment of the point of interest recommendation method of the present invention;

[0066] Figure 5 This is a flowchart illustrating the fourth embodiment of the point-of-interest recommendation method of the present invention.

[0067] Figure 6 This is a structural block diagram of the first embodiment of the point of interest recommendation device of the present invention.

[0068] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0069] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0070] Reference Figure 1 , Figure 1 This is a schematic diagram of the device structure for recommending points of interest in the hardware operating environment involved in the embodiments of the present invention.

[0071] like Figure 1As shown, the point-of-interest recommendation device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen or an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as a disk drive. The memory 1005 may also optionally be a storage device independent of the aforementioned processor 1001.

[0072] Those skilled in the art will understand that Figure 1 The structure shown does not constitute a limitation on the point of interest recommendation device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0073] like Figure 1 As shown, the memory 1005, which serves as a storage medium, may include an operating system, a data storage module, a network communication module, a user interface module, and an interest point recommendation program.

[0074] exist Figure 1 In the point-of-interest (POI) recommendation device shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the POI recommendation device of the present invention can be set in the POI recommendation device, and the POI recommendation device calls the POI recommendation program stored in the memory 1005 through the processor 1001 and executes the POI recommendation method provided in the embodiment of the present invention.

[0075] This invention provides a point-of-interest recommendation method, referring to... Figure 2 , Figure 2 This is a flowchart illustrating the first embodiment of the point of interest recommendation method of the present invention.

[0076] In this embodiment, the point of interest recommendation method includes the following steps:

[0077] Step S10: Obtain candidate points of interest based on the current location of the target terminal;

[0078] It is easy to understand that the execution entity in this embodiment is the server. The server receives the current location sent by the target terminal, and then obtains candidate points of interest based on the target terminal's current location. The target terminal can be understood as the terminal device that the user is using, that is, the terminal device carrying the user terminal. In specific implementation, in order to improve the accuracy of the obtained candidate points of interest, the area where the target terminal is located can be obtained, and the area where the terminal is located can be divided into corresponding area grids through map projection. The target area grid corresponding to the current location of the target terminal is queried in the area grid, and the candidate points of interest corresponding to the target area grid are obtained. The area can be a city, district, county, town, etc., and can be set according to actual needs. This embodiment does not limit this. In specific implementation, when performing map projection on the area where the terminal is located, the Google S2 algorithm can be used to improve spatial indexing efficiency, and further, it also improves the efficiency of locating the target area grid.

[0079] Step S20: Obtain the user's historical behavior data sent by the target terminal, and determine the user preference center words based on the user's historical behavior data of the target terminal;

[0080] It should be noted that after receiving the user's historical behavior data from the target terminal, the server can convert the historical behavior data into corresponding word vectors, calculate the similarity between the word vectors, and then determine the user preference center words based on the similarity. These user preference center words can be understood as the core words representing the current user's behavioral preferences, such as "(preference) hot pot" or "(preference) shopping." The user historical behavior data can be the historical behavior data of the current user stored in the target terminal's search engine logs, such as click logs, search logs, etc., including data on the points of interest clicked by the current user within the same time period over the past 30 days, and the top three search results. Then, semantic analysis is performed on the above user historical behavior data to identify the current user's behavioral preference center words. In a specific implementation, the Continuous Bag-Of-Words (CBOW) model in the word2vec training algorithm can be used to perform semantic analysis on the above user historical behavior data to improve the accuracy of the obtained user preference center words, and further improve the accuracy of obtaining the points of interest to be recommended based on the current user's user preference center words.

[0081] Step S30: Select recommended interest points from the candidate interest points based on the user preference central words, and push the recommended interest points to the target terminal for display.

[0082] It is easy to understand that after determining the current user's user preference core words, the user preference core words can be used to select the points of interest to be recommended from the candidate points of interest, and the points of interest to be recommended can be pushed to the target terminal for display. For example, if the current user's user preference core word is hot pot and the current user is in region A, then the points of interest to be recommended can be popular hot pot restaurants in region A, popular tourist attractions in region A, etc.

[0083] Furthermore, to improve user experience and user engagement, after selecting recommended points of interest from the candidate points of interest, a recommended point of interest list can be generated based on the selected points of interest, and the recommended point of interest list can be pushed to the target terminal for adaptive display. In a specific implementation, the sorting method of the recommended points of interest in the list can also be calculated based on the user's preference keywords. For example, if the current user's preference keyword is hot pot and the current user is in region A, the order of popular hot pot restaurants in region A in the recommended list can be placed before popular tourist attractions in region A, thereby improving user engagement and user experience. In another implementation, a sorting weight value can be set for each recommended point of interest in the recommended list, and then the sorting method of the recommended points of interest can be set according to the sorting weight value. The specific weight value setting method can be set according to actual needs, and this embodiment does not limit it.

[0084] In this embodiment, candidate points of interest (POIs) are obtained based on the current location of the target terminal. User historical behavior data sent by the target terminal is also acquired, and user preference keywords are determined based on this data. POIs to be recommended are then selected from the candidate POIs based on these user preference keywords and pushed to the target terminal for display. Unlike existing technologies that rely solely on user search terms for POI recommendations, resulting in low recommendation relevance, this embodiment improves the accuracy of the acquired POIs by acquiring user historical behavior data sent by the target terminal, determining user preference keywords based on this data, and then selecting POIs to be recommended from the candidate POIs obtained based on the target terminal's current location. This further enhances user engagement and user experience.

[0085] refer to Figure 3 , Figure 3 This is a flowchart illustrating the second embodiment of the point of interest recommendation method of the present invention.

[0086] Based on the first embodiment described above, in this embodiment, step S10 includes:

[0087] Step S101: Obtain the area where the target terminal is located, and divide the area where the terminal is located into corresponding area grids through map projection;

[0088] It should be noted that when obtaining the area where the target terminal is located, the current scene information of the area where the terminal is located can also be obtained. The current scene information can be the current time or the time period of the current time, the road planning of the area where the terminal is located, the activity level of road segments, etc. Then, the corresponding area splitting rules are matched according to the current scene information, and the area where the terminal is located is split into corresponding area grids through map projection according to the area splitting rules. For example, when the current time period is obtained as 1:00 am to 3:00 am, the splitting gradient can be expanded (which can be understood as expanding the area grid obtained by splitting); when the road planning perfection of the area where the terminal is located is less than the preset perfection, the area is split according to the planned roads; for road segments with activity levels lower than the preset activity levels, the area grid corresponding to the road segment can be expanded. The specific degree of grid expansion can be set according to the size of the area where the terminal is located, and this embodiment does not limit this. In specific implementation, the region segmentation rules can be set according to actual needs. For example, the region where the terminal is located can be segmented at equal intervals according to the principle of equal intervals, that is, the obtained region grid can be an equal interval grid; it can also be segmented into grids of different sizes according to the activity level of the current region; it can also be segmented into corresponding irregular grids according to road planning, thereby improving the adaptability of the obtained region grid to the user's region, and further improving the accuracy of the subsequent selection of points of interest based on the region grid.

[0089] Step S102: Query the target area grid corresponding to the current location of the target terminal in the area grid, and obtain the candidate points of interest corresponding to the target area grid.

[0090] It is easy to understand that after the area where the terminal is located is divided into corresponding area grids, the target area grid corresponding to the current position of the target terminal can be queried in the area grid, and the candidate points of interest corresponding to the target area grid can be obtained. The so-called candidate points of interest can be understood as the points of interest of the area grid corresponding to the current position of the target terminal in the area where the terminal is located, that is, the points of interest within a preset range of the target terminal, and the preset range is the target area grid.

[0091] It should be noted that the points of interest (POIs) in the area where the terminal is located can be obtained based on the historical behavior data of the area where the terminal is located. Specifically, regional behavior items can be extracted from the historical behavior data of the area where the terminal is located, and then the number of clicks on the regional behavior items can be counted and sorted to generate a regional behavior ranking result. Based on the regional behavior ranking result, POIs in the area where the terminal is located can be selected from the regional behavior items. Furthermore, the area where the terminal is located can be divided into different regional grids. Then, the target regional grid corresponding to the current location of the target terminal can be queried in the regional grids, and the POIs corresponding to the target regional grid can be found in the POIs of the area where the terminal is located, i.e., the candidate POIs. This improves the efficiency of obtaining candidate POIs and enables the rapid acquisition of popular POIs near the user. Furthermore, it also improves the efficiency of obtaining recommended POIs based on the candidate POIs.

[0092] In this embodiment, the region where the target terminal is located is obtained, and the region is divided into corresponding regional grids through map projection. The target regional grid corresponding to the current location of the target terminal is queried in the regional grid, and the candidate points of interest corresponding to the target regional grid are obtained. By dividing the region where the terminal is located into different regional grids and then querying the target regional grid corresponding to the current location of the target terminal in the regional grid, the efficiency of locating the regional grid where the terminal is located is improved. Furthermore, by searching for the points of interest corresponding to the target regional grid in the points of interest in the region where the terminal is located, the efficiency of obtaining candidate points of interest is improved. This enables the rapid acquisition of popular points of interest near the user, and further improves the efficiency of subsequently obtaining recommended points of interest based on the candidate points of interest.

[0093] refer to Figure 4 , Figure 4 This is a flowchart illustrating the third embodiment of the point of interest recommendation method of the present invention.

[0094] Based on the above embodiments, in this embodiment, step S20 includes:

[0095] Step S201: Obtain the user's historical behavior data sent by the target terminal, and convert the user's historical behavior data into corresponding word vectors;

[0096] It should be noted that when acquiring the user's historical behavior data sent by the target terminal, the user's historical behavior data can first be processed into a text string, then the text string can be segmented to obtain a vocabulary set and a vocabulary sequence. Then, based on the vocabulary sequence, text semantic analysis can be performed on the vocabulary set to determine the central words. Finally, the central words are mapped to a vector space to obtain the corresponding word vectors. In a specific implementation, the Continuous Bag-Of-Words (CBOW) model in the word2vec training algorithm can be used to perform semantic analysis on the vocabulary set to improve the accuracy of the obtained central words, and further improve the accuracy of determining the current user's preference central words based on the central words.

[0097] Step S202: Calculate the similarity between the word vectors and determine the user preference center word based on the similarity.

[0098] It is easy to understand that after obtaining the above word vectors, the similarity between the word vectors can be calculated. Then, the similarity is sorted to obtain a similarity ranking result. Based on the similarity ranking result, user preference center words are selected from the center words. For example, the similarity can be sorted from high to low to obtain a similarity ranking result. Then, the center word corresponding to the word vector with the highest similarity ranking in the similarity ranking result is taken as the user preference center word. This improves the accuracy of obtaining the user preference center words, realizes the accurate extraction of user preferences, and further improves the accuracy of selecting interest points to be recommended from the candidate interest points based on the user preference center words.

[0099] In this embodiment, user historical behavior data sent by the target terminal is acquired and converted into corresponding word vectors. The similarity between the word vectors is calculated, and user preference center words are determined based on the similarity. By performing word segmentation on the user historical behavior data, a vocabulary set is obtained. Then, text semantic analysis is performed on the vocabulary set to determine the center words, thereby improving the accuracy of the obtained center words. The center words are then mapped to a vector space to obtain corresponding word vectors. The similarity between the word vectors is calculated, and user preference center words are determined based on the similarity to improve the accuracy of the obtained user preference center words. This achieves accurate extraction of user preferences and further improves the accuracy of selecting recommended interest points from the candidate interest points based on user preference center words.

[0100] refer to Figure 5 , Figure 5 This is a flowchart illustrating the fourth embodiment of the point of interest recommendation method of the present invention.

[0101] Based on the above embodiments, in this embodiment, before step S10, the method further includes:

[0102] Step S01: Obtain the globally unique identifier of the current user of the target terminal, and determine whether the current user is an unregistered user based on the globally unique identifier;

[0103] Step S02: When the current user is not the unregistered user, perform the step of obtaining the candidate point of interest based on the current location of the target terminal.

[0104] It is easy to understand that before obtaining the current location of the target terminal, the globally unique identifier (GUID) of the current user of the target terminal can be obtained first, and it can be determined whether the current user is an unregistered user based on the globally unique identifier. If the current user is not an unregistered user, the candidate points of interest can be obtained based on the current location of the target terminal, so as to facilitate the subsequent accurate point of interest recommendation service for the current user.

[0105] Step S03: When the current user is the unregistered user, obtain the region where the target terminal is located;

[0106] Step S04: Query the recommended list of regional points of interest corresponding to the region where the terminal is located in the recommended list of regional points of interest, and push the recommended list of regional points of interest to the target terminal for display.

[0107] It should be noted that when the current user is the unregistered user, the region where the target terminal is located can be obtained, and then the regional interest recommendation list corresponding to the region where the terminal is located can be queried in the regional interest recommendation list, and the regional interest recommendation list can be pushed to the target terminal for display.

[0108] In practical implementation, to obtain the recommended list of points of interest for each region, before querying the recommended list of regional points of interest corresponding to the region where the terminal is located, the candidate points of interest for different regions and the ranking factors corresponding to the candidate points of interest can be obtained first. Then, the ranking factors are normalized to obtain normalized ranking factors. The ranking factors can be static scores of points of interest, number of user clicks, number of clicks, click date, etc. In practical implementation, the values ​​and weights of each ranking factor can be obtained, and then the above ranking factors are normalized using the Sigmoid function.

[0109]

[0110] In the formula, x is the value of each ranking factor, and base is the weight of each ranking factor.

[0111] Furthermore, the candidate points of interest in the region are sorted according to the normalized sorting factors to generate a recommended list of regional points of interest for different regions. For example, based on the normalized sorting factors, the normalized sorting factors can be comprehensively sorted by the date of occurrence of each sorting factor and the weight of each normalized sorting factor using the following formula, thereby generating a recommended list of points of interest for each region: score=(∑w*2*(sigmod(parm,base)-0.5))*0.99 dt

[0112] In the formula, score is the score of the candidate interest point in the region, parm is the value of each normalized ranking factor, base is the weight of each ranking factor, dt is the date of occurrence of each ranking factor, and w is the weight of each normalized ranking factor.

[0113] In practice, to obtain regional candidate points of interest for different regions, historical behavior data of different regions can be obtained first. Regional behavior items, such as scenic spots, accommodation and catering, cultural venues, and schools, can be extracted from the historical behavior data. Then, the number of clicks on the regional behavior items can be counted and sorted to generate regional behavior ranking results. Finally, regional candidate points of interest for different regions can be selected from the regional behavior items according to the regional behavior ranking results.

[0114] Furthermore, prior to step S20, the method further includes:

[0115] Step S111: Determine whether the target terminal has received the current user's historical behavior data;

[0116] Step S112: If the user's historical behavior data sent by the target terminal is not received, obtain the region where the target terminal is located;

[0117] Step S04: Query the recommended list of regional points of interest corresponding to the region where the terminal is located in the recommended list of regional points of interest, and push the recommended list of regional points of interest to the target terminal for display.

[0118] It is easy to understand that before obtaining the user's historical behavior data sent by the target terminal, it is also possible to first determine whether the user's historical behavior data of the current user sent by the target terminal has been received. If the user's historical behavior data of the current user sent by the target terminal has been received, then the user preference keywords are determined based on the user's historical behavior data of the target terminal. If the user's historical behavior data of the current user sent by the target terminal has not been received, then the region where the target terminal is located is obtained, the regional interest point recommendation list corresponding to the region where the terminal is located is queried in the regional interest point recommendation list, and the regional interest point recommendation list is pushed to the target terminal for display. The specific method of obtaining the interest point recommendation list is the same as when the current user is not the unregistered user, and will not be described in detail here.

[0119] In this implementation, a globally unique identifier (GUID) of the current user of the target terminal is obtained. Based on this GUID, it is determined whether the current user is an unregistered user. If the current user is an unregistered user, the region where the target terminal is located is obtained, and a regional point of interest (POI) recommendation list corresponding to the terminal's region is retrieved from the POI recommendation list. This POI recommendation list is then pushed to the target terminal for display. Alternatively, if the user's historical behavior data is not received from the target terminal, the region where the target terminal is located is obtained, a regional point of interest recommendation list corresponding to the terminal's region is retrieved from the POI recommendation list, and this POI recommendation list is pushed to the target terminal for display. This allows for relatively accurate POI recommendation services even for unregistered users or target terminals that cannot provide historical behavior data, improving the user experience for unregistered users or terminal users who cannot provide historical behavior data.

[0120] Furthermore, embodiments of the present invention also propose a storage medium storing an interest point recommendation program, wherein the interest point recommendation program, when executed by a processor, implements the steps of the interest point recommendation method as described above.

[0121] Reference Figure 6 , Figure 6 This is a structural block diagram of the first embodiment of the point of interest recommendation device of the present invention.

[0122] like Figure 6 As shown, the point-of-interest recommendation device proposed in this embodiment of the invention includes:

[0123] The candidate point of interest acquisition module 10 is used to acquire candidate points of interest based on the current location of the target terminal;

[0124] User preference acquisition module 20 is used to acquire user historical behavior data sent by the target terminal, and determine user preference core words based on the user historical behavior data of the target terminal;

[0125] The point of interest recommendation module 30 is used to select points of interest to be recommended from the candidate points of interest based on the user preference keywords, and push the points of interest to be recommended to the target terminal for display.

[0126] In this embodiment, candidate points of interest (POIs) are obtained based on the current location of the target terminal. User historical behavior data sent by the target terminal is also acquired, and user preference keywords are determined based on this data. POIs to be recommended are then selected from the candidate POIs based on these user preference keywords and pushed to the target terminal for display. Unlike existing technologies that rely solely on user search terms for POI recommendations, resulting in low recommendation relevance, this embodiment improves the accuracy of the acquired POIs by acquiring user historical behavior data sent by the target terminal, determining user preference keywords based on this data, and then selecting POIs to be recommended from the candidate POIs obtained based on the target terminal's current location. This further enhances user engagement and user experience.

[0127] In addition, the candidate point of interest acquisition module is also used to acquire the area where the target terminal is located, and to divide the area where the terminal is located into corresponding area grids through map projection.

[0128] The candidate point of interest acquisition module 10 is further configured to query the target area grid corresponding to the current location of the target terminal in the area grid, and acquire the candidate point of interest corresponding to the target area grid.

[0129] The candidate point of interest acquisition module 10 is also used to acquire the area where the target terminal is located and the current scene information of the area where the terminal is located;

[0130] The candidate point of interest acquisition module 10 is further configured to match the corresponding region segmentation rule according to the current scene information, and to segment the region where the terminal is located into a corresponding region grid through map projection according to the region segmentation rule.

[0131] The user preference acquisition module 20 is also used to acquire user historical behavior data sent by the target terminal and convert the user historical behavior data into corresponding word vectors;

[0132] The user preference acquisition module 20 is also used to calculate the similarity between the word vectors and determine the user preference center word based on the similarity.

[0133] The user preference acquisition module 20 is also used to acquire user historical behavior data sent by the target terminal, and to perform word segmentation processing on the user historical behavior data to obtain a vocabulary set;

[0134] The user preference acquisition module 20 is further configured to determine the central word based on the vocabulary set and map the central word to a vector space to obtain the corresponding word vector.

[0135] The user preference acquisition module 20 is also used to calculate the similarity between the word vectors and sort the similarity to obtain a similarity ranking result;

[0136] The user preference acquisition module 20 is further configured to select user preference center words from the center words based on the similarity ranking results.

[0137] The candidate point of interest acquisition module 10 is also used to acquire the global unique identifier of the current user of the target terminal, and determine whether the current user is an unregistered user based on the global unique identifier.

[0138] The candidate point of interest acquisition module 10 is further configured to perform the operation of acquiring candidate points of interest based on the current location of the target terminal when the current user is not the unregistered user.

[0139] The point of interest recommendation module 30 is also used to obtain the region where the target terminal is located when the current user is the unregistered user;

[0140] The point of interest recommendation module 30 is also used to query the regional point of interest recommendation list corresponding to the region where the terminal is located in the regional point of interest recommendation list, and push the regional point of interest recommendation list to the target terminal for display.

[0141] The user preference acquisition module 20 is also used to determine whether it has received the current user's historical behavior data sent by the target terminal;

[0142] The point of interest recommendation module 30 is also used to obtain the region where the target terminal is located when the user's historical behavior data sent by the target terminal is not received.

[0143] The point of interest recommendation module 30 is also used to query the regional point of interest recommendation list corresponding to the region where the terminal is located in the regional point of interest recommendation list, and push the regional point of interest recommendation list to the target terminal for display.

[0144] The point of interest recommendation module 30 is also used to obtain regional candidate points of interest in different regions and the ranking factors corresponding to the regional candidate points of interest.

[0145] The point of interest recommendation module 30 is also used to normalize the ranking factor to obtain the normalized ranking factor.

[0146] The point of interest recommendation module 30 is further configured to sort the candidate points of interest in the region according to the normalized sorting factor, and generate a regional point of interest recommendation list for different regions.

[0147] The candidate point of interest acquisition module 10 is also used to acquire regional historical behavior data of different regions and extract regional behavior items from the regional historical behavior data;

[0148] The candidate point of interest acquisition module 10 is also used to count the number of clicks on the regional behavior items, sort the number of clicks, and generate a regional behavior sorting result.

[0149] The candidate interest point acquisition module 10 is further configured to select candidate interest points for different regions from the region behavior items based on the region behavior sorting results.

[0150] The interest point recommendation module 30 is further configured to select interest points to be recommended from the candidate interest points based on the user preference central words, and generate an interest point recommendation list based on the interest points to be recommended;

[0151] The point of interest recommendation module 30 is also used to push the point of interest recommendation list to the target terminal for adaptive display.

[0152] Other embodiments or specific implementations of the point of interest recommendation device of the present invention can be referred to the above-described method embodiments, and will not be repeated here.

[0153] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0154] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0155] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory / random access memory, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0156] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. An interest point recommendation method, characterized in that, The point-of-interest recommendation method includes the following steps: Obtain candidate points of interest based on the current location of the target terminal; Obtain user historical behavior data sent by the target terminal, and determine user preference keywords based on the user historical behavior data of the target terminal; Based on the user preference keywords, select the points of interest to be recommended from the candidate points of interest, and push the points of interest to be recommended to the target terminal for display; The process of obtaining candidate points of interest based on the current location of the target terminal includes: Obtain the target terminal's location and current scene information; Match the corresponding region segmentation rule according to the current scene information, and divide the region where the terminal is located into the corresponding region grid through map projection according to the region segmentation rule; Query the target area grid corresponding to the current location of the target terminal in the area grid, and obtain the candidate points of interest corresponding to the target area grid; The region splitting rules include at least one of the following: when the road planning perfection of the area where the target terminal is located is less than the preset perfection, the region is split according to the planned roads; for road segments with activity levels lower than the preset activity levels, the regional grid corresponding to the road segment is expanded; and the splitting gradient is determined according to the current time.

2. The point-of-interest recommendation method as described in claim 1, characterized in that, The step of obtaining user historical behavior data sent by the target terminal and obtaining user preference keywords based on the user historical behavior data of the target terminal specifically includes: Obtain the user's historical behavior data sent by the target terminal, and convert the user's historical behavior data into corresponding word vectors; Calculate the similarity between the word vectors, and determine the user preference center word based on the similarity.

3. The point-of-interest recommendation method as described in claim 2, characterized in that, The step of acquiring user historical behavior data sent by the target terminal and converting the user historical behavior data into corresponding word vectors specifically includes: The system acquires user historical behavior data sent by the target terminal and performs word segmentation on the user historical behavior data to obtain a vocabulary set. The central word is determined based on the vocabulary set, and the central word is mapped to a vector space to obtain the corresponding word vector.

4. The point-of-interest recommendation method as described in claim 3, characterized in that, The step of calculating the similarity between the word vectors and determining the user preference center word based on the similarity specifically includes: Calculate the similarity between the word vectors, and sort the similarity to obtain the similarity ranking result; User-preferred central words are selected from the central words based on the similarity ranking results.

5. The point-of-interest recommendation method as described in any one of claims 1 to 4, characterized in that, Before the step of obtaining the candidate points of interest based on the current location of the target terminal, the method further includes: Obtain the globally unique identifier of the current user of the target terminal, and determine whether the current user is an unregistered user based on the globally unique identifier; If the current user is not the unregistered user, the step of obtaining the candidate points of interest based on the current location of the target terminal is executed.

6. The point-of-interest recommendation method as described in claim 5, after the step of obtaining the globally unique identifier of the current user of the target terminal and determining whether the current user is an unregistered user based on the globally unique identifier, further includes: When the current user is the unregistered user, obtain the region where the target terminal is located; The system queries the list of recommended regional points of interest (POIs) corresponding to the region where the terminal is located, and pushes the POIs recommendation list to the target terminal for display.

7. The point-of-interest recommendation method as described in any one of claims 1 to 4, further comprising, before the step of obtaining user historical behavior data sent by the target terminal and determining user preference keywords based on the user historical behavior data of the target terminal: Determine whether the target terminal has received the current user's historical behavior data; If the user's historical behavior data is not received from the target terminal, the region where the target terminal is located is obtained. The system queries the list of recommended regional points of interest (POIs) corresponding to the region where the terminal is located, and pushes the POIs recommendation list to the target terminal for display.

8. The point-of-interest recommendation method as described in claim 6 or 7, further comprising, before the step of querying the regional point-of-interest recommendation list corresponding to the region where the terminal is located in the regional point-of-interest recommendation list, and pushing the regional point-of-interest recommendation list to the target terminal for display: Obtain the regional candidate points of interest in different regions and the ranking factors corresponding to the regional candidate points of interest; The sorting factor is normalized to obtain the normalized sorting factor. The candidate interest points in the region are sorted according to the normalized sorting factor to generate a recommended list of regional interest points for different regions.

9. The point-of-interest recommendation method as described in claim 8, further comprising, before the step of obtaining regional candidate points of interest and the ranking factors corresponding to the regional candidate points of interest: Obtain historical behavior data for different regions, and extract regional behavior items from the historical behavior data. The number of clicks on the regional behavior items is counted, and the click counts are sorted to generate a regional behavior ranking result; Based on the ranking results of the regional behaviors, select candidate interest points for different regions from the regional behavior items.

10. The interest point recommendation method as described in any one of claims 1 to 4, wherein the step of selecting interest points to be recommended from the candidate interest points based on the user preference central words, and pushing the interest points to be recommended to the target terminal for display, specifically includes: Based on the user preference keywords, select the interest points to be recommended from the candidate interest points, and generate an interest point recommendation list based on the interest points to be recommended; The list of points of interest recommendations is pushed to the target terminal for adaptive display.

11. An interest point recommendation device, the interest point recommendation device comprising: The candidate point of interest acquisition module is used to acquire candidate points of interest based on the current location of the target terminal; The user preference acquisition module is used to acquire user historical behavior data sent by the target terminal, and determine user preference core words based on the user historical behavior data of the target terminal. The point of interest recommendation module is used to select points of interest to be recommended from the candidate points of interest based on the user preference keywords, and push the points of interest to be recommended to the target terminal for display; The candidate point of interest acquisition module is also used to acquire the area where the target terminal is located and the current scene information; Match the corresponding region segmentation rule according to the current scene information, and divide the region where the terminal is located into the corresponding region grid through map projection according to the region segmentation rule; Query the target area grid corresponding to the current location of the target terminal in the area grid, and obtain the candidate points of interest corresponding to the target area grid; The region splitting rules include at least one of the following: when the road planning perfection of the area where the target terminal is located is less than the preset perfection, the region is split according to the planned roads; for road segments with activity levels lower than the preset activity levels, the regional grid corresponding to the road segment is expanded. Determine the split gradient based on the current time.

12. The point of interest recommendation device as described in claim 11, wherein the user preference acquisition module is further configured to acquire user historical behavior data sent by the target terminal, and convert the user historical behavior data into corresponding word vectors; The user preference acquisition module is also used to calculate the similarity between the word vectors and determine the user preference center word based on the similarity.

13. The interest point recommendation device as described in claim 12, wherein the user preference acquisition module is further configured to acquire user historical behavior data sent by the target terminal, and perform word segmentation processing on the user historical behavior data to obtain a vocabulary set; The user preference acquisition module is further configured to determine the central word based on the vocabulary set and map the central word to a vector space to obtain the corresponding word vector.

14. The interest point recommendation device as described in claim 12, wherein the user preference acquisition module is further configured to calculate the similarity between the word vectors and sort the similarity to obtain a similarity ranking result; The user preference acquisition module is further configured to select user preference center words from the center words based on the similarity ranking results.

15. A point-of-interest recommendation device, characterized in that, The device includes: a memory, a processor, and an interest point recommendation program stored in the memory and executable on the processor, the interest point recommendation program being configured to implement the steps of the interest point recommendation method as described in any one of claims 1 to 10.

16. A storage medium, characterized in that, The storage medium stores an interest point recommendation program, which, when executed by a processor, implements the steps of the interest point recommendation method as described in any one of claims 1 to 10.

Images