Information pushing method, device and computer readable storage medium

Abstract

Embodiments of the present application disclose an information pushing method and device and a computer readable storage medium, which can be applied to various scenes such as cloud technology, AI, intelligent transportation and vehicle-mounted devices. Position information of a target terminal can be collected; a target object classification of a target object associated with the target terminal is obtained, wherein the target object classification is obtained by fusing a static object classification and a dynamic object classification of the target object, and the dynamic object classification is obtained by constructing historical information of the target object; a target label set corresponding to the position information is identified; based on the target object classification and the target label set, to-be-recommended information is determined, and the to-be-recommended information is pushed to the target terminal. In this way, by collecting the position information of the target terminal and obtaining the target object classification closer to the habit characteristics of the user, the to-be-recommended information is determined based on the target object classification and the target label set of the position information, so that the to-be-recommended information is more suitable for the needs of the user, and the accuracy of pushing information is improved.

Description

Technical Field

[0001] This application relates to the field of computer technology, specifically to an information push method, apparatus, and computer-readable storage medium. Background Technology

[0002] With the development of information technology, push notification services have brought users a wealth of experiences. In order to ensure that the pushed information matches the user's needs, related technologies construct user profiles based on relevant data of the user to be recommended to. For example, a corresponding user profile is created based on the user's attribute information, and then information is pushed based on the created user profile.

[0003] In the process of researching and practicing existing technologies, the inventors of this application discovered that when constructing user profiles, existing technologies create profiles based on surface information such as user attribute information. The user profiles constructed are surface-level static profiles, meaning that the constructed user profiles lack potential characteristics. As a result, the information pushed to users has a low degree of relevance to user needs, affecting the accuracy of information push. Summary of the Invention

[0004] This application provides an information push method, apparatus, and computer-readable storage medium. These improvements enhance the accuracy of information push notifications.

[0005] This application provides an information push method, including:

[0006] Collect the location information of the target terminal;

[0007] Obtain a target user profile of the target user associated with the target terminal, wherein the target user profile is obtained by fusing the static user profile and dynamic user profile of the target user, and the dynamic user profile is constructed from the historical displacement information of the target user;

[0008] Identify the set of target tags corresponding to the location information;

[0009] Based on the target user profile and target tag set, the information to be recommended is determined and pushed to the target terminal.

[0010] Accordingly, embodiments of this application provide an information push device, including:

[0011] The acquisition unit is used to acquire the location information of the target terminal;

[0012] The acquisition unit is used to acquire a target user profile of a target user associated with the target terminal, wherein the target user profile is obtained by fusing the static user profile and dynamic user profile of the target user, and the dynamic user profile is constructed from the historical displacement information of the target user;

[0013] The identification unit is used to identify the set of target tags corresponding to the location information;

[0014] The determining unit is used to determine the information to be recommended based on the target user profile and the target tag set, and push the information to be recommended to the target terminal.

[0015] In some embodiments, the information push device further includes a construction unit, configured to:

[0016] Collect target terminal information and corresponding target user information, and generate a static user profile of the target user based on the target terminal information and corresponding target user information;

[0017] Obtain historical displacement information of the target terminal, and determine the dynamic user profile of the target user based on the historical displacement information of the target terminal;

[0018] The static user profile and the dynamic user profile are fused to obtain the target user profile of the target user.

[0019] In some embodiments, the building unit is further configured to:

[0020] Extract displacement data for the target time period from the historical displacement information of the target terminal;

[0021] The location attribute tag set of the target user is determined based on the displacement data of the target time period;

[0022] A dynamic user profile of the target user is constructed based on the set of location attribute tags.

[0023] In some embodiments, the building unit is further configured to:

[0024] The displacement data for the target time period is converted into a displacement trajectory image;

[0025] Extract a target trajectory sub-image from the displacement trajectory image that conforms to a preset trajectory point density rule;

[0026] The trajectory point data in the target trajectory sub-image is classified using a preset classification decision tree to obtain the location attribute label corresponding to the target trajectory sub-image;

[0027] The location attribute tag set of the target user is determined based on the location attribute tags.

[0028] In some embodiments, the building unit is further configured to:

[0029] Identify displacement trajectory point data in the displacement trajectory image of the target time period;

[0030] Cluster the displacement trajectory point data to obtain clustered displacement trajectory point data;

[0031] The clustered displacement trajectory point data is sampled to obtain the target trajectory point data;

[0032] Generate a corresponding trajectory buffer image based on the target trajectory point data, and extract a target trajectory sub-image from the trajectory buffer image that conforms to a preset trajectory point density rule.

[0033] In some embodiments, the building unit is further configured to:

[0034] The trajectory buffer image is subjected to trajectory point recognition to obtain trajectory point recognition results;

[0035] Based on the trajectory point recognition results, determine the target image region in the trajectory buffer image that conforms to the preset trajectory point density rule;

[0036] The target image region is extracted from the trajectory buffer image to obtain the target trajectory sub-image.

[0037] In some embodiments, the building unit is further configured to:

[0038] Obtain the probability value of each location attribute tag in the location attribute tag set;

[0039] When the probability value of the detected location attribute label is less than a preset label probability threshold, the location attribute label that is less than the preset label probability threshold is determined as a location label to be determined, and the target location corresponding to the location label to be determined is determined.

[0040] Extract the set of environmental images corresponding to the target location from the historical displacement information of the target terminal; the set of environmental images contains multiple environmental images.

[0041] Determine the target location attribute label corresponding to the target location based on the multiple environmental images;

[0042] The target location attribute tag set is updated using the target location attribute tag set, and a dynamic user profile of the target user is constructed based on the updated location attribute tag set.

[0043] In some embodiments, the building unit is further configured to:

[0044] Perform image semantic recognition on the plurality of environmental images to obtain the target environment identifiers corresponding to the plurality of environmental images;

[0045] The target location attribute label corresponding to the target location is determined based on the target environment identifier.

[0046] Furthermore, this application also provides a computer device, including a processor and a memory, wherein the memory stores an application program, and the processor is used to run the application program in the memory to implement the steps in the information push method provided in this application.

[0047] Furthermore, embodiments of this application also provide a computer-readable storage medium storing a plurality of instructions adapted for loading by a processor to execute steps in any of the information push methods provided in embodiments of this application.

[0048] Furthermore, embodiments of this application also provide a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps in any of the information push methods provided in embodiments of this application.

[0049] This application embodiment can be applied to various scenarios such as cloud technology, AI, smart transportation, and vehicle-mounted systems. Specifically, it can collect the location information of the target terminal; obtain a target user profile of the target user associated with the target terminal, wherein the target user profile is obtained by fusing the static user profile and dynamic user profile of the target user, and the dynamic user profile is constructed from the historical displacement information of the target user; identify the target tag set corresponding to the location information; determine the information to be recommended based on the target user profile and the target tag set, and push the information to be recommended to the target terminal. Thus, this application embodiment collects the location information of the target terminal in real time to facilitate subsequent information push based on the real-time location information; and obtains the target user profile of the target user, which, since it is obtained by fusing the static user profile and dynamic user profile, makes the target user profile closer to the user's habit characteristics; furthermore, by combining the target user profile and the location information of the target terminal, the information to be pushed is determined and pushed to the user's terminal, so that the pushed information is more in line with the user's needs and the accuracy of information push is improved. Attached Figure Description

[0050] 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.

[0051] Figure 1 This is a schematic diagram of a scenario for the information push system provided in an embodiment of this application;

[0052] Figure 2 This is a flowchart illustrating the steps of the information push method provided in the embodiments of this application;

[0053] Figure 3 This is a flowchart illustrating another step of the information push method provided in this application embodiment;

[0054] Figure 4 This is a block flowchart illustrating the dynamic user profile construction method provided in this application embodiment;

[0055] Figure 5 This is a schematic diagram of the structure of the information push device provided in the embodiments of this application;

[0056] Figure 6 This is a schematic diagram of the structure of the computer device provided in the embodiments of this application. Detailed Implementation

[0057] 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.

[0058] This application provides an information push method, apparatus, and computer-readable storage medium. Specifically, this application will describe the information push apparatus from the perspective of the information push device, which can be integrated into a computer device, such as a server or a terminal. The server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, etc., but is not limited to these. The terminal and server can be directly or indirectly connected via wired or wireless communication, which is not limited herein.

[0059] The solutions provided in this application involve technologies such as artificial intelligence information push, which are specifically illustrated through the following embodiments:

[0060] For example, see Figure 1This is a schematic diagram of an information push system provided in an embodiment of this application. This scenario can be applied to various scenarios such as cloud technology, AI, smart transportation, and vehicle-mounted systems, including a terminal 10 and a server 20. The terminal 10 and the server 20 are connected wirelessly to achieve data interaction.

[0061] Terminal 10 or the target application on terminal 10 can be used to obtain location information in real time. This location information may include trajectory point information (geographical location or address information) for each location, environmental image information corresponding to each trajectory point, and displacement trajectory information composed of each trajectory point. Terminal 10 sends this location information to server 20; it is also used to obtain target terminal information and corresponding target user information from terminal 10, and send the target terminal information and corresponding target user information to server 20; so that server 20 can perform corresponding data processing, such as generating a target user profile, determining information to be recommended, and pushing information.

[0062] Server 20 can collect the location information of the target terminal (terminal 10); obtain the target user profile of the target user associated with the target terminal, wherein the target user profile is obtained by fusing the static user profile and dynamic user profile of the target user, and the dynamic user profile is constructed by the historical displacement information of the target user; identify the target tag set corresponding to the location information; determine the information to be recommended based on the target user profile and the target tag set, and push the information to be recommended to the target terminal, that is, return to terminal 10.

[0063] Information push can include processing methods such as generating target user profiles, collecting location information of target terminals, obtaining corresponding target user profiles, determining information to be recommended, and sending the information to be recommended.

[0064] The following sections provide detailed descriptions of each example. It should be noted that the order of the following embodiments is not intended to limit the preferred order of the embodiments.

[0065] In this embodiment, the description will focus on an information push device, which can be integrated into a computer device such as a terminal or server. See also Figure 2 , Figure 2 This application provides a flowchart illustrating the steps of an information push method. Taking an example where the information push device is specifically integrated into a server, the specific process when the processor on the server executes the program corresponding to the information processing method is as follows:

[0066] 101. Collect the location information of the target terminal.

[0067] The target terminal can be a mobile terminal or a terminal device installed on some carrier. This terminal can have the function of recording displacement information, including the surrounding environment information of a fixed location and the displacement trajectory information during movement. Furthermore, the target terminal can also have communication capabilities, enabling data interaction with a server. Specifically, the target terminal can use its communication function to send the recorded surrounding environment information of the fixed location and / or the displacement trajectory information during movement to the server. It should be noted that a dedicated target application for interacting with the server can also be installed on the target terminal, allowing the terminal to send the acquired information data to the server.

[0068] For example, taking a car's dashcam as the target terminal, the dashcam acquires the user's location information, surrounding environment information, and displacement trajectory information in real time during driving and / or stopping, and sends this information to the server in real time. It should be noted that a dedicated application for interacting with the server can also be installed on the target terminal, allowing the dashcam to send the data to the server.

[0069] This location information can be the physical address information recorded by the target terminal at the current moment, reflecting the user's geographical location and travel information in real time. For example, assuming the target application on the target terminal is running, it can record the user's geographical location in real time by running the application in the background. This allows the terminal to query and utilize the location information later, or send the recorded location information to a server for further processing.

[0070] To ensure the effectiveness of the pushed information and its compliance with user needs, this application embodiment can determine the recommended information to be pushed by combining location information. Specifically, this application embodiment can determine the recommended information by combining the user's real-time location information. Collecting the target terminal's location information may include receiving location information sent by the target terminal. This facilitates the subsequent determination of the recommended information to be pushed based on the target terminal's location information, ensuring the effectiveness of the subsequently pushed recommended information, such as ensuring that the recommended information (e.g., information about offline milk tea shops) matches the target terminal's current location information and meets the user's needs.

[0071] In some implementations, the step "collecting the location information of the target terminal" may further include: obtaining the target account information of the target user associated with the target terminal; obtaining historical displacement information corresponding to the target account information from the displacement information database; and extracting the location information corresponding to the target time from the historical displacement information. Additionally, it may include: obtaining historical displacement information corresponding to the identifier of the target terminal from the displacement information database, and extracting the location information corresponding to the target time from the historical displacement information. It should be noted that the target time mentioned above can be the previous second, the previous 10 seconds, or the previous minute, etc.

[0072] By using the above methods, the location information of the target terminal can be obtained, so as to determine the recommended information to be pushed in combination with the location information of the target terminal, and ensure the effectiveness of the recommended information, such as ensuring that the recommended information (such as offline milk tea store information) can match the current location information of the target terminal and meet the needs of the user.

[0073] 102. Obtain the target user profile of the target user associated with the target terminal.

[0074] The target user profile can be a user information model constructed based on the user's relevant data. Specifically, the target user profile is obtained by fusing the target user's static user profile and dynamic user profile. The static user profile can be generated from the target terminal information and the user information of the target user corresponding to the target terminal, while the dynamic user profile can be generated from the target user's displacement information (such as historical displacement information).

[0075] To obtain a target user profile for a target user corresponding to a target terminal, embodiments of this application can obtain the corresponding target user profile based on the target user's target account information. Specifically, the target user profile can be obtained by: determining the target account information of the corresponding target user based on the identifier of the target terminal, and matching the target user profile corresponding to the target account information from a preset profile database, wherein the preset profile database contains the correspondence between account information and user profiles. Furthermore, embodiments of this application can also obtain the target user profile of the corresponding target user based on the identifier of the target terminal.

[0076] In some implementations, the step of "obtaining the target user profile of the target user associated with the target terminal" may include:

[0077] (1) Collect target terminal information and corresponding target user information, and generate a static user profile of the target user based on the target terminal information and corresponding target user information.

[0078] The target terminal information may include attribute information of the target terminal and / or the carrier on which the target terminal is located, such as the brand, product model, logo, shape, and configuration of the target terminal, and / or the brand, product signal, logo, configuration, and power of the carrier on which the target terminal is located. For example, taking an in-vehicle device as the target terminal, such as a dashcam, the target terminal information may include the brand, product model, logo, shape, and configuration of the in-vehicle device, as well as the brand, product signal, logo, configuration, power, and engine displacement of the vehicle.

[0079] The target user information can be the basic information of the end user, such as the user's gender, age, occupation, education level, etc.

[0080] The static user profile can be a user information model constructed based on static data such as terminal information and user information, and it can include terminal information and basic information of the terminal user.

[0081] To generate a static user profile for a target user, this embodiment of the application first needs to collect target terminal information and corresponding target user information, and then generate a static user profile based on the target terminal information and corresponding target user information. Specifically, target terminal information corresponding to the target terminal identifier can be obtained from a static information database, as well as pre-stored target user information can be obtained from the static information database. It should be noted that when a user registers through a target terminal, they can fill in the target terminal information and their personal target user information, which are then stored in the server's static information database for subsequent data retrieval. Furthermore, the obtained target terminal information and corresponding target user information are integrated to construct a static user profile for the target user. This allows the static user profile to be used for information recommendation or other purposes.

[0082] (2) Obtain the historical displacement information of the target terminal and determine the dynamic user profile of the target user based on the historical displacement information of the target terminal.

[0083] The historical displacement information can be displacement data of the target terminal within any historical time period. This historical displacement information can reflect the displacement trajectory and locations visited by the user corresponding to the target terminal within any historical time period. The historical displacement information can include displacement trajectory data of any historical time period, location information at historical moments, and corresponding environmental information.

[0084] The dynamic user profile can be a user information model constructed from dynamic data of end users, such as dynamic data such as end user operation data and displacement data. For example, in the embodiments of the application, a user profile can be constructed based on displacement trajectory data, which can reflect the user's habitual preferences for a certain geographical location and their habitual activity preferences in that geographical location.

[0085] To obtain a dynamic user profile of the target user, this embodiment of the application needs to obtain the historical displacement information of the target terminal in advance. For example, the historical displacement information of the target terminal can be obtained by retrieving the historical displacement information corresponding to the identifier of the target terminal from a preset displacement information database. Then, the dynamic user profile of the target user is determined based on the historical displacement information of the target terminal. In this way, dynamic data mining of the terminal user (the target user in this embodiment of the application) is realized, and a dynamic user profile is constructed so that the dynamic user profile can be used for information push or other purposes in the future.

[0086] (3) The static user profile and the dynamic user profile are merged to obtain the target user profile of the target user.

[0087] To obtain a target user profile for the target user, this embodiment of the application, after obtaining a static user profile and a dynamic user profile, can fuse the static user profile and the dynamic user profile to obtain the target user profile for the target user. In this way, the target user profile, which includes both the static and dynamic user profiles, can be used to determine subsequent information to be recommended, improving the accuracy of subsequent information recommendations to the user.

[0088] In some implementations, the step "determine the dynamic user profile of the target user based on the historical displacement information of the target terminal" may include:

[0089] (2.1) Extract displacement data for the target time period from the historical displacement information of the target terminal.

[0090] The displacement data can be trajectory data generated by the end user during displacement, which may include multiple trajectory point data. Each trajectory point data corresponds to a specific moment in the displacement process, meaning that one trajectory point data can be generated at each moment. For example, during the end user's journey, the target terminal acquires the target user's trajectory point data in real time and sends the trajectory point data to the server in real time. This allows the server to concatenate any adjacent trajectory point data based on the timestamp of each trajectory point data to obtain the user's displacement trajectory data over historical time.

[0091] The target time period can be a preset time period within a historical period. In this embodiment, the target time period can represent the time period involved in studying the activity habits or preferences of end users (related users). When studying and building user profiles, displacement data corresponding to the target time period can be selected. The target time period can be the past day, week, month, etc., and this embodiment does not impose specific limitations.

[0092] It should be noted that, due to the large amount of historical displacement data generated by each target terminal, using all displacement data corresponding to the target terminal as research data for studying and building user profiles would not only consume a large amount of server system resources, such as memory, but also require a significant time cost. Furthermore, some of the historical displacement data generated by the target terminal may contain data from 5 or 10 years ago. Over time, the user's habitual preferences for a particular geographical location and their habitual activities in that location may change. If all historical displacement data is used as research data for studying and building user profiles, the resulting user profiles may not align with the user's habits and preferences, leading to lower accuracy in information delivery and impacting user experience.

[0093] To obtain the research data (displacement data) corresponding to the expected results, this application embodiment needs to extract displacement data for the target time period from the historical displacement information of the target terminal. This target time period can be 1 day ago, 1 week ago, 1 month ago, 3 months ago, or 1 year ago, etc. Specifically, after obtaining the historical displacement information of the target terminal, the displacement data corresponding to the target time period is extracted from the historical displacement information. This facilitates subsequent research on the dynamic aspects of the user profile based on the displacement data of the target time period. For example, it allows for research on the geographical location habits and preferences of the target users on the target terminal, as well as their activity habits and preferences in that geographical location, to construct a corresponding dynamic user profile. This makes the information pushed later more aligned with the target users' habits and preferences, improving the accuracy of subsequent information pushes.

[0094] (2.2) Determine the set of location attribute tags for the target user based on the displacement data of the target time period.

[0095] The location attribute label set is a dataset containing one or more location attribute sub-labels. These sub-labels can be attribute labels for a specific geographic location, reflecting the target user's habits and / or preferences regarding that location or area. For example, if the target user of a target terminal frequently spends a significant amount of time in a residential area, the location attribute sub-label for that residential area relative to the target user can be determined to be "home" or "residential," or similar labels. Similarly, if the target user of a target terminal frequently stays in or is active near a shopping mall, the location attribute sub-label for that shopping mall relative to the target user can be determined to be "leisure" or "shopping," etc.

[0096] In order to obtain the set of location attribute tags of the target user, this embodiment of the application needs to determine all the location attribute sub-tags of the target user based on the displacement data of the target time period. Then, the set of location attribute tags of the target user is determined based on all the location attribute sub-tags. In this way, the user's permanent geographical location can be known based on the set of location attribute tags, so as to study the geographical location of the target user and the activity habits and preferences of the target user in the geographical location. This enables research on the dynamic data of the target user, so as to facilitate the subsequent construction of a dynamic user profile of the target user.

[0097] In some implementations, the step "determine the set of location attribute tags for the target user based on the displacement data for the target time period" may include:

[0098] (2.2.1) Convert the displacement data of the target time period into a displacement trajectory image.

[0099] The displacement trajectory image can be an image containing the displacement trajectory pattern, that is, the displacement trajectory is displayed in the form of an image. Through the displacement trajectory image, the distribution of each displacement trajectory point can be understood.

[0100] To obtain a displacement trajectory image containing displacement data, this embodiment of the application converts the displacement data for the target time period into a displacement trajectory image after obtaining the displacement data. Specifically, it identifies multiple trajectory point data contained in the displacement data for the target time period; obtains the address information and timestamp of each trajectory point data; extracts a target sub-image from a preset map based on the address information of each trajectory point data; marks the address information of each trajectory point data as a trajectory point in the target sub-image to obtain a target trajectory point image containing multiple trajectory points; and establishes trajectory routes between any two adjacent trajectory points in the target trajectory point image based on the chronological order of the timestamps in each trajectory point data to obtain a displacement trajectory image. This displacement trajectory image contains the continuous displacement trajectory of the target user within the target time period, i.e., the displacement trajectories in the displacement trajectory image are obtained by sequentially connecting the trajectory points according to their corresponding timestamp order via trajectory routes.

[0101] In addition, displacement data can be converted into displacement trajectory images using data image conversion tools, which is not limited here.

[0102] The displacement data for the target time period is converted into displacement trajectory images through the above methods. This allows us to view the displacement trajectory in image form, understand the distribution of each trajectory point, and use it for subsequent user profile construction.

[0103] (2.2.2) Extract the target trajectory sub-image from the displacement trajectory image that conforms to the preset trajectory point density rule.

[0104] The preset trajectory point density rule can be a constraint rule for trajectory point density, used to select image regions that meet the trajectory point density requirements as trajectory sub-images. For example, the preset trajectory point density rule can be to select the target image region with the highest density in the displacement trajectory image, or it can be to select the image regions corresponding to the top three larger densities in the displacement trajectory image, or the preset trajectory point density rule can be to select image regions with a number greater than the preset trajectory point number threshold; no specific limitation is made here.

[0105] The target trajectory sub-image can be an image where the distance between two adjacent trajectory points in the displacement trajectory is small or they overlap, meaning that the density of trajectory points in the target trajectory sub-image is high and the number of trajectory points is large. This target trajectory sub-image can reflect the time a target user of the target terminal spends at a specific geographical location. Therefore, the information contained in the target trajectory sub-image can be used as research data to understand the target user's habitual preferences for geographical locations and their habitual activity preferences at those locations, thereby constructing a corresponding dynamic user profile.

[0106] To obtain target trajectory sub-images as research data, this embodiment of the application, after obtaining the displacement trajectory image corresponding to the target time period, can extract target trajectory sub-images that conform to a preset trajectory point density rule from the displacement trajectory image. Specifically, trajectory points in the displacement trajectory image can be identified to obtain trajectory point identification results. Based on the trajectory point identification results, the image region in the displacement trajectory image that conforms to the preset trajectory point density rule is cropped to obtain the corresponding target trajectory sub-image. It should be noted that this embodiment of the application can obtain one or more target trajectory sub-images from the displacement trajectory image. In this way, the target trajectory sub-image can be used as research data to understand the target user's habitual preferences for geographical location and their habitual activity preferences in that geographical location, thereby constructing a corresponding dynamic user profile.

[0107] The step "extracting a target trajectory sub-image from the displacement trajectory image that conforms to a preset trajectory point density rule" may include:

[0108] (2.2.2.1) Identify displacement trajectory point data in the displacement trajectory image of the target time period;

[0109] (2.2.2.2) Cluster the displacement trajectory point data to obtain the clustered displacement trajectory point data;

[0110] (2.2.2.3) Sample the clustered displacement trajectory point data to obtain the target trajectory point data;

[0111] (2.2.2.4) Generate a corresponding trajectory buffer image based on the target trajectory point data, and extract the target trajectory sub-image that conforms to the preset trajectory point density rule from the trajectory buffer image.

[0112] It should be noted that displacement trajectory images can contain high-precision trajectory data. When generating high-precision trajectory data, the interval between each trajectory point can be one second, that is, one trajectory point is generated per second. Therefore, the amount of data corresponding to high-precision trajectory data is relatively large.

[0113] To obtain concise trajectory data from displacement trajectory images, this embodiment of the application performs noise reduction and compression processing on the trajectory data in the displacement trajectory images. Specifically, displacement trajectory point data in the displacement trajectory images of the target time period are identified, and the displacement trajectory point data is clustered to filter out noisy and invalid data, resulting in filtered displacement trajectory point data, i.e., clustered displacement trajectory point data. Further, the clustered displacement trajectory point data is compressed. For example, the clustered displacement trajectory point data is sampled. The sampling method can be uniform sampling according to a preset sampling rule, such as sampling every 5 trajectory points or every 10 trajectory points; this is not limited here. Alternatively, the sampling method can be Douglas-Pukherfae method. Through the above sampling methods, sampled target trajectory point data can be obtained. It should be noted that after sampling, multiple target trajectory point data are obtained, and this embodiment of the application does not limit the specific number of sampled target trajectory point data.

[0114] It should be noted that when the target terminal generates trajectory point data, factors such as network signal latency, weak signal, and lag may result in multiple trajectory point data with errors. After these erroneous trajectory point data are uploaded to the server, the server processes the uploaded trajectory point data to obtain a displacement trajectory image. Part of the displacement trajectory corresponding to the erroneous trajectory point data may deviate from the road in the displacement trajectory image, such as this part of the displacement trajectory deviating from the track in the image and appearing inside a building. Furthermore, factors causing partial displacement trajectory deviation may also include clustering and sampling processing of the displacement trajectory point data. Because clustering and sampling processes remove some trajectory point data, and when generating the corresponding displacement trajectory based on the clustered and sampled target trajectory point data, adjacent trajectory points are connected by straight paths, which may cause some displacement trajectories in the subsequent displacement trajectory image to deviate from the road in the displacement trajectory image.

[0115] To ensure full utilization of all target trajectory point data after sampling, this embodiment generates a corresponding trajectory buffer image based on all the target trajectory point data after obtaining the sampled target trajectory point data. This ensures that the trajectory point corresponding to each target trajectory point data and the connected trajectory path have a corresponding buffer area in the image, preventing the discarding of parts of the displacement trajectory image that deviate from the path in the displacement trajectory image, thus ensuring full utilization of the target trajectory point data.

[0116] Furthermore, in order to obtain target trajectory sub-images as research data, embodiments of this application can acquire target trajectory sub-images that conform to preset trajectory point density rules from displacement trajectory images, such as cropping image regions that conform to preset trajectory point density rules in displacement trajectory images to obtain corresponding target trajectory sub-images. It should be noted that embodiments of this application can acquire one or more target trajectory sub-images from displacement trajectory images.

[0117] In some implementations, the step "extracting a target trajectory sub-image that conforms to a preset trajectory point density rule from the trajectory buffer image" may include: performing trajectory point recognition on the trajectory buffer image to obtain trajectory point recognition results; determining the target image region in the trajectory buffer image that conforms to the preset trajectory point density rule based on the trajectory point recognition results; and extracting the target image region from the trajectory buffer image to obtain the target trajectory sub-image.

[0118] Specifically, trajectory point recognition can be performed on the trajectory buffer image using a preset image recognition tool or a preset analysis function to obtain trajectory point recognition results. These results can include density information (such as density values) of the trajectory points contained in each image region of the trajectory buffer image. For example, trajectory point recognition can be performed on the trajectory buffer image using a kernel density estimation function to obtain the results. Further, target image regions in the trajectory buffer image that conform to preset trajectory point density rules are determined based on the trajectory point recognition results. For example, the trajectory point density values ​​of each image region in the trajectory buffer image are obtained from the trajectory point recognition results, and the trajectory point density values ​​of each image region are compared with preset trajectory point density thresholds in the preset trajectory point density rules. Image regions with trajectory point density values ​​greater than the preset trajectory point density thresholds are determined as target image regions. Finally, the determined target image regions in the trajectory buffer image are cropped to obtain target trajectory sub-images corresponding to each target image region.

[0119] Therefore, the target trajectory sub-image can be used as research data to understand the target terminal's target users' habitual preferences for geographical locations and their habitual preferences for activities in those geographical locations, thereby constructing a corresponding dynamic user profile.

[0120] (2.2.3) The trajectory point data in the target trajectory sub-image is classified by a preset classification decision tree to obtain the position attribute label corresponding to the target trajectory sub-image.

[0121] It should be noted that this preset classification decision tree can be constructed based on the location attribute label (Point of Interest, POI) data of open source websites. This preset classification decision tree is used to identify and classify the location attribute labels of each geographic location or region.

[0122] To obtain the location attribute labels corresponding to the target trajectory sub-image, this embodiment of the application, after obtaining the target image, can classify the trajectory point data in the target trajectory sub-image using a preset classification decision tree to obtain the location attribute labels corresponding to the target trajectory sub-image. Specifically, multiple trajectory point data contained in the target trajectory sub-image are obtained; each preset classification decision tree is used for classification processing to obtain the location attribute sub-label corresponding to each trajectory point data, that is, multiple location attribute sub-labels corresponding to the target trajectory sub-image are obtained; the location attribute labels corresponding to the target trajectory sub-image are determined based on the multiple location attribute sub-labels. For example, the multiple location attribute sub-labels are classified by attribute to obtain the location attribute sub-label corresponding to each attribute category; the ratio between the number of attribute sub-labels corresponding to each attribute category and the total number of multiple location attribute sub-labels is calculated to obtain the probability value corresponding to each attribute category; the location attribute sub-label corresponding to the attribute category with the highest probability value is determined as the location attribute label corresponding to the target trajectory sub-image. In this way, the location attribute labels corresponding to the target trajectory sub-image can be used to construct a dynamic user profile of the target user of the target terminal.

[0123] (2.2.4) Determine the set of location attribute tags for the target user based on the location attribute tags.

[0124] The location attribute label set is a dataset containing one or more location attribute sub-labels. These location attribute sub-labels can be attribute labels for a specific geographic location, reflecting the target user's habits and / or preferences for that geographic location or region.

[0125] It should be noted that in the displacement trajectory image corresponding to the displacement data of the target time period, multiple target trajectory sub-images that conform to the preset trajectory point density rules can be extracted. The trajectory point data in each target trajectory sub-image is classified by a preset classification decision tree to obtain the position attribute label corresponding to each target trajectory sub-image. At this time, multiple position attribute labels corresponding to the displacement trajectory image can be obtained.

[0126] To obtain the set of location attribute labels corresponding to the displacement data for the target time period, this embodiment of the application, after obtaining the location attribute labels corresponding to each target trajectory sub-image, generates a set of location attribute labels for the target user based on all the location attribute labels, thus obtaining the set of location attribute labels corresponding to the displacement data for the target time period. In this way, the user's permanent geographical location can be determined based on the set of location attribute labels, allowing for research on the target user's geographical location and activity habits and preferences in that location. This enables research based on the target user's dynamic data, facilitating the subsequent construction of a dynamic user profile for the target user.

[0127] (2.3) Construct a dynamic user profile of the target user based on the set of location attribute tags.

[0128] The dynamic user profile can be a user information model constructed from dynamic data of end users, such as dynamic data such as the displacement data of end users, which reflects the user's habitual preferences for a certain geographical location and the habitual preferences for activities in that geographical location.

[0129] To obtain a dynamic user profile of a target user, this embodiment of the application, after obtaining the set of location attribute tags for the target user, can construct a dynamic user profile based on the set of location attribute tags. This allows for the construction of a corresponding dynamic user profile based on the target user's geographical location habits and preferences, as well as their activity habits and preferences in that geographical location, so that the dynamic user profile can be used for subsequent information push or other purposes.

[0130] In some implementations, the step "constructing a dynamic user profile of the target user based on the set of location attribute tags" may include:

[0131] (2.3.1) Obtain the probability value of each location attribute label in the location attribute label set.

[0132] Here, the probability value represents the probability that the geographic area corresponding to the target trajectory sub-image belongs to the location attribute label. It should be noted that in a real-world geographic area, such as a residential community, this community may include a kindergarten, several offline stores, and numerous residential buildings. The probability values ​​for "kindergarten," "offline stores," and "residential buildings" in this geographic area can be calculated separately, depending on the number of buildings; no specific limit is imposed here. It is understandable that since the primary use of this geographic area is residential, the probability value for "residential buildings" is much greater than the probability values ​​for "kindergarten" and "offline stores." Therefore, the location attribute label corresponding to this geographic area is "home," not "school" or "education" as for "kindergarten," and not "leisure" or "commercial" as for "offline stores."

[0133] It should be noted that each position attribute tag in the set of position attribute tags has a corresponding probability value. Specifically, since each position attribute tag can be determined by multiple position attribute sub-tags, the probability value of each position attribute tag can be the probability value of a certain position attribute sub-tag among the multiple position attribute sub-tags, that is, this probability value has been determined when the position attribute tag is determined.

[0134] In order to ensure the accuracy of each location attribute tag in the location attribute tag set, this application embodiment needs to obtain the probability value of each location attribute tag in the location attribute tag set, so that it can be used to determine whether the corresponding location attribute tag corresponds to the actual preferred geographical location of the target user associated with the target terminal, thereby improving the accuracy of the subsequent generation of dynamic user profiles.

[0135] (2.3.2) When the probability value of the detected location attribute label is less than the preset label probability threshold, the location attribute label that is less than the preset label probability threshold is determined as the location label to be determined, and the target location corresponding to the location label to be determined is determined.

[0136] The preset label probability threshold is used to filter location attribute labels that need to be reconfirmed. For example, when the probability value of a detected location attribute label is greater than or equal to the preset label probability threshold, the location attribute labels corresponding to the preset label probability threshold can be directly used to construct a dynamic user profile; when the probability value of a detected location attribute label is less than the preset label probability threshold, it indicates that the accuracy of the location attribute label that is less than the preset label probability threshold is biased, and the location attribute labels of the corresponding geographical locations need to be further reconfirmed.

[0137] For example, the preset tag probability threshold is set to 90%. When the probability value of a certain location attribute tag is greater than or equal to 90%, the location attribute tag is determined to be accurate and can be used to build a dynamic user profile. When the probability value of a certain location attribute tag is less than 90%, the location attribute tag is determined as a location tag to be determined, and the target location corresponding to the location tag to be determined is determined, so as to further determine the location attribute tag of the target location.

[0138] (2.3.3) Extract the set of environmental images corresponding to the target location from the historical displacement information of the target terminal. The set of environmental images contains multiple environmental images.

[0139] The environmental image can be an image of the surrounding environment of the target location. This environmental image is automatically acquired by the target terminal at the target location through planar acquisition of the surrounding environment. After acquisition, the target terminal will upload the environmental image and the corresponding trajectory point data to the server for storage in real time.

[0140] (2.3.4) Determine the target location attribute label corresponding to the target location based on multiple environmental images.

[0141] It should be noted that the environmental image contains environmental information surrounding the target location, such as road signs, billboards, street views, buildings, and markings on buildings (such as names, icons, signs, etc.). Based on the environmental information contained in the environmental image, the activity habits and preferences of the target user at the target location can be determined. Therefore, the location attribute label of the target user at the target location can be determined based on the environmental information contained in the environmental image.

[0142] For example, in the set of location attribute labels, a location attribute label with a probability less than a preset label threshold is "office," indicating that the target location corresponding to this location attribute label mainly belongs to an office area, which may include a large number of office buildings. However, this office area may also include some leisure areas, such as milk tea shops, convenience stores, and supermarkets. To further determine the target user's habitual preferences for the target location area and their activity habits in that area, it is necessary to make determinations based on multiple environmental images corresponding to the target location. For example, the environmental information in the environmental images collected by the target terminal includes milk tea shops and milk tea shop markers, indicating that the target user's actual location attribute label in the target location area is "leisure," not "office." Therefore, "leisure" can be confirmed as the target location attribute label corresponding to the target user in that target location.

[0143] In some implementations, the step "determine the target location attribute label corresponding to the target location information based on multiple environmental images" may include: performing image semantic recognition on multiple environmental images to obtain target environment identifiers corresponding to multiple environmental images; and determining the target location attribute label corresponding to the target location based on the target environment identifiers.

[0144] Specifically, multiple environmental images can be semantically recognized using a pre-set recognition model to obtain target environment identifiers corresponding to the multiple environmental images. This pre-set recognition model can be a pre-trained image-text recognition model. Furthermore, the target location attribute label of the target user at the target location is determined based on the target environment identifiers corresponding to the multiple environmental images.

[0145] (2.3.5) Update the set of location attribute tags using the target location attribute tags, and construct a dynamic user profile of the target user based on the updated set of location attribute tags.

[0146] To obtain an accurate set of location attribute tags, this embodiment of the application, after redetermining the target location attribute tags, replaces the location attribute tags in the set that are less than a preset tag probability threshold with the newly determined target location attribute tags, thereby updating the location attribute tag set and obtaining an updated location attribute tag set. Then, a dynamic user profile of the target user is constructed based on the location attribute tags contained in the updated location attribute tag set.

[0147] By using the above methods, a dynamic user profile can be constructed based on the target user's geographical location habits and preferences and activity habits in that geographical location. This allows the dynamic user profile to be fused with the static user profile in the embodiments of this application to obtain the target user profile. Based on this target user profile, the information that the target user is interested in can be determined, thereby improving the accuracy of subsequent information recommendations.

[0148] 103. Identify the set of target labels corresponding to the location information.

[0149] The location information refers to the physical address information recorded by the target terminal at the current moment. It reflects the terminal user's geographical location and travel information in real time. It should be noted that this location information is not limited to the target user's specific current location; it can also be used to identify a preset area surrounding the target user's current location. For example, it could represent a commercial area, residential area, administrative area, or office area within 3 kilometers of the target user's current location. No specific limitations are imposed here.

[0150] The target tag set can contain one or more target location attribute sub-tags, representing the specific location or location attribute of the current location information. For example, if the location area corresponding to the current location information includes office area, leisure area, and commercial area, then the corresponding target tag set can include target location attribute sub-tags such as office, leisure, and commercial. It should be noted that target location attribute sub-tags can also be further refined attribute tags, such as milk tea shop, supermarket, shopping mall, office building, and amusement park, etc., without limitation here.

[0151] It should be noted that before determining the recommended information to be pushed, the target location attribute sub-tags contained in the current location can be used to determine the services that the target user's current location can provide; and then, the recommended information can be determined based on one or more target location attribute sub-tags contained in the current location and the target user profile.

[0152] To determine the services available at a target user's current location, this embodiment of the application can extract an image of the target location region corresponding to the location information according to a preset range rule, and classify the data in the target location region image using a preset classification decision tree to obtain a target tag set corresponding to the target location region image. This target tag set contains multiple target location attribute sub-tags. Therefore, based on the multiple target location attribute sub-tags contained in the target tag set, the services available at the current location can be determined, facilitating the subsequent determination of recommendation information based on the services provided at the current location and the target user profile.

[0153] 104. Based on the target user profile and target tag set, determine the information to be recommended and push the information to be recommended to the target terminal.

[0154] The information to be recommended can be information that the target user is interested in. For example, the information to be recommended could be discount information for a product, information about famous stores selling a product, articles related to a profession, educational resources, brand information of the target device, software update information of the target device, etc., which is determined based on the target user profile and the current location information of the target device.

[0155] To obtain the recommended information for a target user, this embodiment of the application, after acquiring the target user profile and the current location information of the target terminal, determines the recommended information for the target user based on the target tag set corresponding to the target user profile and the current location information. Specifically, it acquires one or more target location attribute sub-tags contained in the target tag set; matches the acquired target location attribute sub-tags with the location attribute tags contained in the target user profile; determines the target location attribute sub-tags that match the corresponding location attribute tags in the target user profile as recommended attribute tags; searches for the information corresponding to the recommended attribute tag, and determines the information corresponding to the recommended attribute tag as the recommended information.

[0156] In some implementations, after the step "matching the acquired target location attribute sub-tags with the location attribute tags contained in the target user profile", the process may include: when multiple target location attribute sub-tags matching the corresponding location attribute tags in the target user profile are detected, acquiring the target time corresponding to the location information, and determining the target location attribute sub-tags matching the corresponding location attribute tags in the target user profile as candidate attribute tags, thus obtaining multiple candidate attribute tags; acquiring the time corresponding to each location attribute tag contained in the target user profile, and determining a recommended attribute tag based on the time corresponding to each location attribute tag, the target time, and the multiple candidate attribute tags; searching for the information corresponding to the recommended attribute tag, and determining the information corresponding to the recommended attribute tag as the information to be recommended. Thus, when the current location area of ​​the target user contains multiple target location attribute sub-tags matching the target user profile, the time of each location attribute tag contained in the target user profile is acquired to understand the target user's historical preferences for the corresponding geographical location and activities in that geographical location. Then, combined with the current target time, the recommended attribute tags of the target user in the current location information and target time are determined, so as to determine recommended information based on the recommended attribute tags and push the recommended information, thereby improving the accuracy of information recommendation.

[0157] Furthermore, when multiple target location attribute sub-tags are detected that match the corresponding location attribute tags in the target user profile, each matching sub-tag can be identified as a recommended attribute tag, resulting in multiple recommended attribute tags. Then, recommended information corresponding to each recommended attribute tag is retrieved and pushed to the target user's target terminal simultaneously or sequentially. This allows for the delivery of multiple recommended information messages for the target user to choose from, improving the user experience.

[0158] In this embodiment, to make the user profile more closely match the habits and characteristics of end users, a target user profile is pre-built by combining static and dynamic user profiles. Specifically, firstly, a static user profile of the target user is generated based on the target terminal information and the corresponding target user information; then, the historical displacement information of the target terminal is obtained, and displacement data for the target time period is extracted from the historical displacement information to determine the target user's location attribute tag set based on the displacement data for the target time period. Furthermore, a corresponding target user profile is constructed based on the target user's location attribute tag set, thereby enabling the determination of the user's permanent geographical location and the construction of a dynamic user profile; finally, the static and dynamic user profiles are merged to obtain the merged target user profile. This allows for subsequent understanding of the target user's habitual preferences for the target geographical location and their activity habits within that location, facilitating the determination of the information to be recommended.

[0159] It should be noted that, to avoid changes in user habits and preferences over time, which could lead to a target user profile that doesn't accurately reflect recent activity habits, this application embodiment selects displacement data from the historical displacement information corresponding to the target terminal during the target time period as research data when constructing the dynamic user profile. This ensures that the subsequently studied and constructed user profile is more closely aligned with the user's recent activity habits and preferences. Furthermore, one or more target trajectory sub-images with high-density trajectory points are identified from the displacement data during the target time period, and the trajectory point data of these target trajectory sub-images are classified using a classification decision tree to determine the corresponding location attribute labels. In addition, a dynamic user profile corresponding to the target user is constructed based on the determined location attribute labels. Through the above methods, the user's permanent geographical location is obtained through the determined location attribute labels, allowing for the study of the target user's geographical location and activity habits and preferences within that location. This enables research based on the target user's dynamic data, thus constructing a dynamic user profile for the target user.

[0160] Furthermore, based on the target user profile and the real-time location information of the target terminal, the system determines the information that the target user is interested in and pushes it to the target user's target terminal, so that the pushed information is more in line with the user's needs and the accuracy of the pushed information is improved.

[0161] As can be seen from the above, the embodiments of this application can be applied to various scenarios such as cloud technology, AI, smart transportation, and vehicle in-vehicle systems. Specifically, it can collect the location information of the target terminal; obtain the target user profile of the target user associated with the target terminal, wherein the target user profile is obtained by fusing the static user profile and dynamic user profile of the target user, and the dynamic user profile is constructed from the historical displacement information of the target user; identify the target tag set corresponding to the location information; determine the information to be recommended based on the target user profile and the target tag set, and push the information to be recommended to the target terminal. Thus, the embodiments of this application collect the location information of the target terminal in real time to facilitate subsequent information push based on the real-time location information; and obtain the target user profile of the target user, which, since it is obtained by fusing the static user profile and dynamic user profile, makes the target user profile closer to the user's habit characteristics; furthermore, by combining the target user profile and the location information of the target terminal, the information to be pushed is determined and pushed to the user's terminal, so that the pushed information is more in line with the user's needs and the accuracy of the pushed information is improved.

[0162] Based on the method described in the above embodiments, the following examples will provide further detailed explanations.

[0163] The embodiments of this application can be applied to various scenarios such as cloud technology, AI, smart transportation, and vehicle systems. Specifically, this application takes an information push device as an example to further describe the information push method provided by this application embodiment.

[0164] See Figure 3 , Figure 3 This is a schematic flowchart of another step in the information push method provided in the embodiments of this application. Figure 4 This is a block flowchart illustrating the dynamic user profile construction method provided in this application embodiment; for ease of understanding, please refer to it in conjunction with... Figure 3 and Figure 4 The embodiments of this application will be described below.

[0165] In this embodiment, the description will focus on an information push device, which can be integrated into a computer device such as a server. When the processor on the server executes the program corresponding to the information push method, the specific process of the information push method is as follows:

[0166] 201. Collect target terminal information and corresponding target user information, and generate a static user profile of the target user based on the target terminal information and corresponding target user information.

[0167] The target terminal can be a mobile terminal or a terminal device installed on some carrier. This terminal can record displacement information, including the surrounding environment of a fixed location and the displacement trajectory during movement. Furthermore, the target terminal can also have communication capabilities, enabling data interaction with a server. Specifically, the target terminal can use its communication function to send the recorded surrounding environment information and / or displacement trajectory information to the server. It should be noted that a dedicated application for interacting with the server can also be installed on the target terminal to send the information data acquired by the terminal to the server. For example, using a car's dashcam as the target terminal, the dashcam can acquire the user's location information, surrounding environment information, and displacement trajectory information in real time during driving and / or while stationary, and send this information to the server in real time. Again, a dedicated application for interacting with the server can be installed on the target terminal to send the information data acquired by the dashcam to the server.

[0168] The target terminal information may include attribute information of the target terminal and / or the carrier on which the target terminal is located, such as the brand, product model, logo, shape, and configuration of the target terminal, and / or the brand, product signal, logo, configuration, and power of the carrier on which the target terminal is located. For example, taking an in-vehicle device as the target terminal, such as a dashcam, the target terminal information may include the brand, product model, logo, shape, and configuration of the in-vehicle device, as well as the brand, product signal, logo, configuration, power, and engine displacement of the vehicle.

[0169] The target user information can be the basic information of the end user, such as the user's gender, age, occupation, education level, etc.

[0170] The static user profile can be a user information model constructed based on static data such as terminal information and user information, and it can include terminal information and basic information of the terminal user.

[0171] To generate a static user profile for the target user, this embodiment of the application obtains target terminal information corresponding to the target terminal identifier from a static information database, and obtains pre-stored target user information from the static information database. It should be noted that when a user registers through a target terminal, they can fill in the target terminal information and their personal target user information, which are then stored in the static information database of the server for subsequent data retrieval. Furthermore, the obtained target terminal information and corresponding target user information are integrated to construct a static user profile for the target user.

[0172] 202. Obtain the historical displacement information of the target terminal, and determine the dynamic user profile of the target user based on the historical displacement information of the target terminal.

[0173] The historical displacement information can be displacement data of the target terminal within any historical time period. This historical displacement information can reflect the displacement trajectory and locations visited by the user corresponding to the target terminal within any historical time period. The historical displacement information can include displacement trajectory data of any historical time period, location information at historical moments, and corresponding environmental information.

[0174] The dynamic user profile can be a user information model constructed from dynamic data of end users, such as dynamic data such as end user operation data and displacement data. For example, in the embodiments of the application, a user profile can be constructed based on displacement trajectory data, which can reflect the user's habitual preferences for a certain geographical location and their habitual activity preferences in that geographical location.

[0175] In order to construct a dynamic user profile of the target user, this application embodiment extracts displacement data of the target time period from the historical displacement information of the target terminal, determines the location attribute tag set of the target user based on the displacement data of the target time period, and constructs a dynamic user profile of the target user based on the location attribute tag set.

[0176] Specifically, the process of determining the set of location attribute labels for the target user may include: converting displacement data of the target time period into displacement trajectory images; identifying displacement trajectory point data in the displacement trajectory images of the target time period; clustering the displacement trajectory point data to obtain clustered displacement trajectory point data; sampling the clustered displacement trajectory point data to obtain target trajectory point data; generating a corresponding trajectory buffer image based on the target trajectory point data; performing trajectory point recognition on the trajectory buffer image to obtain trajectory point recognition results; determining the target image region in the trajectory buffer image that conforms to a preset trajectory point density rule based on the trajectory point recognition results; extracting the target image region from the trajectory buffer image to obtain a target trajectory sub-image; classifying the trajectory point data in the target trajectory sub-image using a preset classification decision tree to obtain the location attribute labels corresponding to the target trajectory sub-image; and determining the set of location attribute labels for the target user based on the location attribute labels.

[0177] For example, taking a smart in-vehicle device (such as a dashcam) as the target terminal, this device can collect displacement trajectory information, such as trajectory point data, in real time and upload the collected trajectory point data to the server. The server then constructs a dynamic user profile based on the received and stored displacement trajectory data. Specifically, the server obtains the displacement data for the target time period based on the identification information of the smart in-vehicle device. For the obtained high-precision trajectory data (one point per second on the map, resulting in a large amount of data), preprocessing (denoising and compression) is required. For example, the DBSCAN clustering algorithm can be used to cluster the displacement data for the target time period to remove noise and invalid data. Furthermore, uniform sampling and / or the Douglas-Puk method can be used to sample the clustered displacement data for compression, resulting in processed target trajectory point data. Then, ArcGIS tools are used to generate trajectory buffer image features corresponding to the target trajectory point data, and kernel density analysis is used to perform trajectory point density analysis on the trajectory buffer image. The analysis results are then used to extract the target trajectory sub-images that conform to the preset trajectory point density rules from the trajectory buffer image. Next, the trajectory point data in the target trajectory sub-image is classified using a pre-established attribute label classification decision tree to obtain the location attribute labels corresponding to the target trajectory sub-image. In building this attribute label classification decision tree, attribute label (Point of Interest, POI) data obtained from existing map data on open-source websites is combined with national classification standards for POIs to construct attribute label classification decision trees for each location area. Finally, a dynamic user profile of the target user is constructed based on the determined set of location attribute labels.

[0178] It should be noted that, in order to further improve the accuracy of constructing dynamic user profiles and make them more closely reflect the travel habits of target users, location attribute tags with probability values ​​lower than a preset tag probability threshold need further verification. Specifically, the target locations corresponding to location attribute tags with probability values ​​lower than the preset tag probability threshold are obtained. Since the area corresponding to this target location belongs to a region with ambiguous location attribute tags, the location attribute tags in this region need to be re-verified. For example, an environmental image set corresponding to the target location is extracted from the historical displacement information of the target terminal. A pre-trained image semantic segmentation model (Fully Convolutional Network (FCN) and SegNet method) is used to perform semantic segmentation on multiple environmental images in the set, yielding image recognition results. These environmental images contain environmental information surrounding the target location, such as road signs, billboards, street views, buildings, and markings on buildings (e.g., names, icons, signs, etc.). SegNet is a fully convolutional neural network consisting of an encoder and decoder, representing a highly efficient pixel-level semantic segmentation framework. Based on this image recognition result, the target location attribute label is redefined. This allows for further refinement of the POI type attribute and specific address resolution for the preferred region, accurately determining the target user's behavioral characteristics at the target location, such as "leisure—drinking milk tea, shopping, etc." Finally, the target location attribute label set is updated using this target location attribute label, resulting in an updated location attribute label set. This updated location attribute label set is then used to construct a dynamic user profile of the target user.

[0179] 203. Merge the static user profile and the dynamic user profile to obtain the target user profile of the target user.

[0180] In order to obtain the target user profile of the target user, this application embodiment can merge the static user profile and the dynamic user profile after obtaining the static user profile and the dynamic user profile to obtain the target user profile of the target user.

[0181] By using the above methods, static user profiles and dynamic user profiles can be merged to obtain target user profiles. This allows for the provision of personalized recommendation services to users based on the target user profiles and real-time location information, thereby improving the accuracy of information recommendations to users.

[0182] 204. Collect the location information of the target terminal.

[0183] This location information can be the physical address information recorded by the target terminal at the current moment, reflecting the user's geographical location and travel information in real time. For example, assuming the target application on the target terminal is running, it can record the user's geographical location in real time by running the application in the background. This allows the terminal to query and utilize the location information later, or send the recorded location information to a server for further processing.

[0184] To ensure the effectiveness of the pushed information and its compliance with user needs, this application embodiment can determine the recommended information to be pushed by combining location information. Collecting the location information of the target terminal may include: obtaining the target user's target account information from the target terminal; obtaining historical displacement information corresponding to the target account information from a displacement information database; and extracting the location information corresponding to the target time from the historical displacement information. Furthermore, it may also include: obtaining historical displacement information corresponding to the target terminal's identifier from the displacement information database, and extracting the location information corresponding to the target time from the historical displacement information. It should be noted that the target time can be the previous second, the previous 10 seconds, or the previous minute, etc. This facilitates the subsequent determination of the recommended information to be pushed based on the target terminal's location information, ensuring the effectiveness of the subsequently pushed recommended information, such as ensuring that the recommended information (e.g., information about offline milk tea shops) matches the target terminal's current location information and meets the user's needs.

[0185] 205. Obtain the target user profile of the target user associated with the target terminal.

[0186] The target user profile can be a user information model constructed based on the user's relevant data. Specifically, the target user profile is obtained by fusing the target user's static user profile and dynamic user profile. The static user profile can be generated from the target terminal information and the user information of the target user corresponding to the target terminal, while the dynamic user profile can be generated from the target user's displacement information (such as historical displacement information).

[0187] To obtain a target user profile for a target user corresponding to a target terminal, embodiments of this application can obtain the corresponding target user profile based on the target user's target account information. Specifically, the target user profile can be obtained by: determining the target account information of the corresponding target user based on the identifier of the target terminal, and matching the target user profile corresponding to the target account information from a preset profile database, wherein the preset profile database contains the correspondence between account information and user profiles. Furthermore, embodiments of this application can also obtain the target user profile of the corresponding target user based on the identifier of the target terminal.

[0188] 206. Based on the target user profile and the location information of the target terminal, determine the information to be recommended.

[0189] The information to be recommended can be information that the target user is interested in. For example, the information to be recommended could be discount information for a product, information about famous stores selling a product, articles related to a profession, educational resources, brand information of the target device, software update information of the target device, etc., which is determined based on the target user profile and the current location information of the target device.

[0190] In order to obtain the recommendation information for the target user, this application embodiment identifies the target tag set corresponding to the location information after obtaining the location information and the target user profile, and determines the recommendation information based on the target user profile and the target tag set.

[0191] Specifically, the target location area image corresponding to the location information is extracted according to preset range rules, and the data in the target location area image is classified through a preset classification decision tree to obtain a target label set corresponding to the target location area image. The target label set contains multiple target location attribute sub-labels. The obtained target location attribute sub-labels are matched with the location attribute labels contained in the target user profile. The target location attribute sub-labels that match the corresponding location attribute labels in the target user profile are determined as recommended attribute labels. The information corresponding to the recommended attribute label is found and the information corresponding to the recommended attribute label is determined as the information to be recommended.

[0192] By employing the methods described above and leveraging the tags of their target user profiles, service providers can offer personalized recommendations based on user characteristics. Specifically, location tags reveal a user's place of residence, workplace, and shopping locations. Furthermore, detailed attribute tag information can be derived through geocoding of latitude and longitude coordinates. Based on the location of different users at different times, service providers can offer more granular personalized recommendations and decision support, enhancing users' sense of belonging to the products and services and improving the user experience.

[0193] 207. Push the recommended information to the target terminal.

[0194] By executing steps 201-207, a static user profile of the vehicle driver can be constructed based on the binding of the human-vehicle relationship, combining the attributes of the in-vehicle intelligent device with the basic attributes of the vehicle driver. Then, a dynamic user profile of the driver can be constructed based on historical displacement data. The static user profile and the dynamic user profile are merged to obtain the target user profile, so that personalized recommendation services can be provided to the user when using relevant application products, based on the target user profile and real-time location information.

[0195] like Figure 4The diagram shown is a block flowchart illustrating the dynamic user profile construction method provided in this application embodiment. In this application embodiment, steps 201-203 are executed to achieve the following: Figure 4 The process, specifically... Figure 4 The specific process of building a dynamic user profile is as follows:

[0196] 301. Obtain the displacement data of the target terminal during the target time period; perform DBSCAN clustering algorithm on the displacement data to cluster the displacement data during the target time period in order to remove noise and invalid data; and perform sampling on the clustered displacement data through uniform sampling and / or Douglas-Puk method to achieve compression processing, so as to obtain the processed target trajectory point data.

[0197] 302. Use ArcGIS tools to generate trajectory buffer image features corresponding to the target trajectory point data, and use kernel density analysis function to perform trajectory point density analysis on the trajectory buffer image. Extract the target trajectory sub-image from the trajectory buffer image that conforms to the preset trajectory point density rules based on the analysis results. Then, classify the trajectory point data in the target trajectory sub-image using a pre-established attribute label classification decision tree to obtain the position attribute labels corresponding to the target trajectory sub-image, and construct the corresponding position attribute label set.

[0198] 303. For location attribute labels in the location attribute label set whose probability value is less than the preset label probability threshold, the target location corresponding to the attribute label is determined as the location attribute label fuzzy region; by extracting images from the video stream of the historical displacement information uploaded by the target terminal, the set of environmental images corresponding to the target location is obtained; then, based on the fully convolutional network (FCN) in the SegNet framework, multiple environmental images in the environmental image set are semantically segmented to obtain the image recognition result; based on the image recognition result and the target location information (such as latitude and longitude, address, etc.), the specific target location attribute label is determined.

[0199] 304. Update the set of location attribute tags using the target location attribute tags to obtain the updated set of location attribute tags.

[0200] 305. Construct a dynamic user profile of the target user using the updated set of location attribute tags.

[0201] Through the above process, after obtaining a dynamic user profile, travel-related characteristics can be segmented and identified to determine the travel behavior and preference characteristics of target users. This allows for precise location targeting of individual drivers based on travel purpose characteristics and location tags within the dynamic user profile, thereby enabling the segmentation of driver user groups. This can be achieved by categorizing drivers based on travel purpose characteristics (e.g., commuters, primarily for commercial purposes, others) or by location tags (e.g., primarily for work, avid shoppers, etc.). Furthermore, user resources within groups with similar travel profiles can be integrated and utilized to achieve precise recommendations among similar groups.

[0202] As can be seen from the above, the embodiments of this application can be applied to various scenarios such as cloud technology, AI, smart transportation, and vehicle in-vehicle systems. Specifically, it can collect the location information of the target terminal; obtain the target user profile of the target user associated with the target terminal, wherein the target user profile is obtained by fusing the static user profile and dynamic user profile of the target user, and the dynamic user profile is constructed from the historical displacement information of the target user; identify the target tag set corresponding to the location information; determine the information to be recommended based on the target user profile and the target tag set, and push the information to be recommended to the target terminal. Thus, the embodiments of this application collect the location information of the target terminal in real time to facilitate subsequent information push based on the real-time location information; and obtain the target user profile of the target user, which, since it is obtained by fusing the static user profile and dynamic user profile, makes the target user profile closer to the user's habit characteristics; furthermore, by combining the target user profile and the location information of the target terminal, the information to be pushed is determined and pushed to the user's terminal, so that the pushed information is more in line with the user's needs and the accuracy of the pushed information is improved.

[0203] To better implement the above methods, this application also provides an information push device, which can be integrated into network devices, such as servers or terminals. The terminal may include tablet computers, laptop computers, and / or personal computers.

[0204] For example, such as Figure 5 As shown, the information push device may include a collection unit 501, an acquisition unit 502, an identification unit 503, and a determination unit 504.

[0205] The acquisition unit 501 is used to acquire the location information of the target terminal;

[0206] The acquisition unit 502 is used to acquire the target user profile of the target user associated with the target terminal. The target user profile is obtained by fusing the static user profile and dynamic user profile of the target user. The dynamic user profile is constructed from the historical displacement information of the target user.

[0207] The identification unit 503 is used to identify the target tag set corresponding to the location information;

[0208] The determining unit 504 is used to determine the information to be recommended based on the target user profile and the target tag set, and push the information to be recommended to the target terminal.

[0209] In some embodiments, the information push device further includes a construction unit, specifically configured to: collect target terminal information and corresponding target user information, and generate a static user profile of the target user based on the target terminal information and corresponding target user information; obtain historical displacement information of the target terminal, and determine a dynamic user profile of the target user based on the historical displacement information of the target terminal; and fuse the static user profile and the dynamic user profile to obtain the target user profile of the target user.

[0210] In some embodiments, the construction unit is further configured to: extract displacement data for a target time period from the historical displacement information of the target terminal; determine a set of location attribute tags for the target user based on the displacement data for the target time period; and construct a dynamic user profile of the target user based on the set of location attribute tags.

[0211] In some embodiments, the construction unit is further configured to: convert displacement data of the target time period into displacement trajectory images; extract target trajectory sub-images that conform to preset trajectory point density rules from the displacement trajectory images; classify the trajectory point data in the target trajectory sub-images using a preset classification decision tree to obtain the location attribute labels corresponding to the target trajectory sub-images; and determine the location attribute label set of the target user based on the location attribute labels.

[0212] In some embodiments, the construction unit is further configured to: identify displacement trajectory point data in the displacement trajectory image of the target time period; cluster the displacement trajectory point data to obtain clustered displacement trajectory point data; sample the clustered displacement trajectory point data to obtain target trajectory point data; generate a corresponding trajectory buffer image based on the target trajectory point data, and extract a target trajectory sub-image that conforms to a preset trajectory point density rule from the trajectory buffer image.

[0213] In some embodiments, the construction unit is further configured to: perform trajectory point recognition on the trajectory buffer image to obtain trajectory point recognition results; determine the target image region in the trajectory buffer image that conforms to a preset trajectory point density rule based on the trajectory point recognition results; and extract the target image region from the trajectory buffer image to obtain a target trajectory sub-image.

[0214] In some embodiments, the construction unit is further configured to: obtain the probability value of each location attribute tag in the location attribute tag set; when the probability value of a location attribute tag is detected to be less than a preset tag probability threshold, determine the location attribute tag less than the preset tag probability threshold as a location tag to be determined, and determine the target location corresponding to the location tag to be determined; extract the set of environmental images corresponding to the target location from the historical displacement information of the target terminal, the set of environmental images containing multiple environmental images; determine the target location attribute tag corresponding to the target location based on the multiple environmental images; update the location attribute tag set using the target location attribute tag, and construct a dynamic user profile of the target user based on the updated location attribute tag set.

[0215] In some embodiments, the construction unit is further configured to: perform image semantic recognition on multiple environmental images to obtain target environment identifiers corresponding to the multiple environmental images; and determine the target location attribute label corresponding to the target location based on the target environment identifiers.

[0216] In some embodiments, the determining unit 504 is further configured to: obtain one or more location attribute sub-labels contained in the area corresponding to the current location information based on the location information of the target terminal; match the obtained location attribute sub-labels with the location attribute labels contained in the target user profile; when a location attribute sub-label that matches the corresponding location attribute label in the target user profile is detected, determine the location attribute sub-label that matches the corresponding location attribute label in the target user profile as a recommended attribute label; search for the information corresponding to the recommended attribute label, and determine the information corresponding to the recommended attribute label as information to be recommended.

[0217] As can be seen from the above, the embodiments of this application can be applied to various scenarios such as cloud technology, AI, smart transportation, and vehicle in-vehicle systems. Specifically, the location information of the target terminal can be collected by the collection unit 501; the target user profile of the target user associated with the target terminal can be obtained by the acquisition unit 502, wherein the target user profile is obtained by fusing the static user profile and dynamic user profile of the target user, and the dynamic user profile is constructed by the historical displacement information of the target user; the target tag set corresponding to the location information can be identified by the identification unit 503; and the target user profile to be recommended can be determined by the determination unit 504 based on the target user profile and the target tag set, and the recommended information can be pushed to the target terminal. Thus, the embodiments of this application collect the location information of the target terminal in real time, so as to facilitate subsequent information push based on the real-time location information; and obtain the target user profile of the target user. Since the target user profile is obtained by fusing the static user profile and the dynamic user profile, the target user profile can be closer to the user's habit characteristics; then, by combining the target user profile and the location information of the target terminal, the information to be pushed can be determined and pushed to the user's terminal, so that the pushed information is more in line with the user's needs and the accuracy of the pushed information can be improved.

[0218] This application also provides a computer device, such as... Figure 6 As shown, it illustrates a structural schematic diagram of the computer device involved in the embodiments of this application, specifically:

[0219] The computer device may include components such as a processor 601 with one or more processing cores, a memory 602 with one or more computer-readable storage media, a power supply 603, and an input unit 604. Those skilled in the art will understand that... Figure 6 The computer device structure shown does not constitute a limitation on the computer device and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein:

[0220] The processor 601 is the control center of the computer device. It connects various parts of the computer device via various interfaces and lines. By running or executing software programs and / or modules stored in the memory 602, and by calling data stored in the memory 602, it performs various functions of the computer device and processes data, thereby performing overall detection of the computer device. Optionally, the processor 601 may include one or more processing cores; preferably, the processor 601 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may also not be integrated into the processor 601.

[0221] The memory 602 can be used to store software programs and modules. The processor 601 executes various functional applications and data processing by running the software programs and modules 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 computer device, etc. In addition, the memory 602 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 602 may also include a memory controller to provide the processor 601 with access to the memory 602.

[0222] The computer device also includes a power supply 603 that supplies power to the various components. Preferably, the power supply 603 can be logically connected to the processor 601 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply 603 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0223] The computer device may also include an input unit 604, which can be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0224] Although not shown, the computer device may also include a display unit, etc., which will not be described in detail here. Specifically, in the embodiments of this application, the processor 601 in the computer device loads the executable files corresponding to the processes of one or more applications into the memory 602 according to the following instructions, and the processor 601 runs the applications stored in the memory 602 to realize various functions, as follows:

[0225] The system collects the location information of the target terminal; obtains the target user profile of the target user associated with the target terminal, wherein the target user profile is obtained by fusing the static user profile and dynamic user profile of the target user, and the dynamic user profile is constructed from the historical displacement information of the target user; identifies the target tag set corresponding to the location information; and determines the information to be recommended based on the target user profile and the target tag set, and pushes the information to be recommended to the target terminal.

[0226] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0227] As can be seen from the above, the embodiments of this application can be applied to various scenarios such as cloud technology, AI, smart transportation, and vehicle in-vehicle systems. Specifically, it can collect the location information of the target terminal; obtain the target user profile of the target user associated with the target terminal, wherein the target user profile is obtained by fusing the static user profile and dynamic user profile of the target user, and the dynamic user profile is constructed from the historical displacement information of the target user; identify the target tag set corresponding to the location information; determine the information to be recommended based on the target user profile and the target tag set, and push the information to be recommended to the target terminal. Thus, the embodiments of this application collect the location information of the target terminal in real time to facilitate subsequent information push based on the real-time location information; and obtain the target user profile of the target user, which, since it is obtained by fusing the static user profile and dynamic user profile, makes the target user profile closer to the user's habit characteristics; furthermore, by combining the target user profile and the location information of the target terminal, the information to be pushed is determined and pushed to the user's terminal, so that the pushed information is more in line with the user's needs and the accuracy of the pushed information is improved.

[0228] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0229] Therefore, embodiments of this application provide a computer-readable storage medium storing a plurality of instructions that can be loaded by a processor to execute steps in any of the information push methods provided in embodiments of this application. For example, the instructions can execute the following steps:

[0230] The system collects the location information of the target terminal; obtains the target user profile of the target user associated with the target terminal, wherein the target user profile is obtained by fusing the static user profile and dynamic user profile of the target user, and the dynamic user profile is constructed from the historical displacement information of the target user; identifies the target tag set corresponding to the location information; and determines the information to be recommended based on the target user profile and the target tag set, and pushes the information to be recommended to the target terminal.

[0231] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0232] The computer-readable storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0233] Since the instructions stored in the computer-readable storage medium can execute the steps in any of the information push methods provided in the embodiments of this application, the beneficial effects that any of the information push methods provided in the embodiments of this application can achieve can be realized. For details, please refer to the previous embodiments, which will not be repeated here.

[0234] The above provides a detailed description of an information push method, apparatus, and computer-readable storage medium provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is 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. An information push method, characterized in that, include: Collect the location information of the target terminal; Obtain a target user profile of the target user associated with the target terminal, wherein the target user profile is obtained by fusing the static user profile and dynamic user profile of the target user, and the dynamic user profile is constructed from the historical displacement information of the target user; wherein the historical displacement information is high-precision global positioning system trajectory data collected by the vehicle-mounted device; The process of constructing the dynamic user profile includes: Extract displacement data for the target time period from the historical displacement information; The displacement data for the target time period is converted into a displacement trajectory image; Extract a target trajectory sub-image that conforms to a preset trajectory point density rule from the displacement trajectory image, wherein the preset trajectory point density rule is used to identify and extract dense trajectory point regions corresponding to the target user's dwell behavior from the displacement trajectory image, as the target trajectory sub-image; The trajectory point data in the target trajectory sub-image is classified by a preset classification decision tree to obtain the set of location attribute labels corresponding to the target trajectory sub-image. The preset classification decision tree is pre-constructed based on the point of interest classification standard of open source map data and is used to perform semantic classification of the geographical area where the trajectory points are located. The dynamic user profile is constructed based on the set of location attribute tags; Identify the set of target tags corresponding to the location information; Based on the target user profile and target tag set, the information to be recommended is determined and pushed to the target terminal.

2. The method according to claim 1, characterized in that, Before obtaining the target user profile of the target user associated with the target terminal, the process includes: Collect target terminal information and corresponding target user information, and generate a static user profile of the target user based on the target terminal information and corresponding target user information; Obtain historical displacement information of the target terminal, and determine the dynamic user profile of the target user based on the historical displacement information of the target terminal; The static user profile and the dynamic user profile are fused to obtain the target user profile of the target user.

3. The method according to claim 1, characterized in that, Extracting a target trajectory sub-image from the displacement trajectory image that conforms to a preset trajectory point density rule includes: Identify displacement trajectory point data in the displacement trajectory image of the target time period; Cluster the displacement trajectory point data to obtain clustered displacement trajectory point data; The clustered displacement trajectory point data is sampled to obtain the target trajectory point data; Generate a corresponding trajectory buffer image based on the target trajectory point data, and extract a target trajectory sub-image from the trajectory buffer image that conforms to a preset trajectory point density rule.

4. The method according to claim 3, characterized in that, Extracting a target trajectory sub-image from the trajectory buffer image that conforms to a preset trajectory point density rule includes: The trajectory buffer image is subjected to trajectory point recognition to obtain trajectory point recognition results; Based on the trajectory point recognition results, determine the target image region in the trajectory buffer image that conforms to the preset trajectory point density rule; The target image region is extracted from the trajectory buffer image to obtain the target trajectory sub-image.

5. The method according to claim 1, characterized in that, The step of constructing a dynamic user profile of the target user based on the location attribute tag set includes: Obtain the probability value of each location attribute tag in the location attribute tag set; When the probability value of the detected location attribute label is less than a preset label probability threshold, the location attribute label that is less than the preset label probability threshold is determined as a location label to be determined, and the target location corresponding to the location label to be determined is determined. Extract the set of environmental images corresponding to the target location from the historical displacement information of the target terminal; the set of environmental images contains multiple environmental images. Determine the target location attribute label corresponding to the target location based on the multiple environmental images; The target location attribute tag set is updated using the target location attribute tag set, and a dynamic user profile of the target user is constructed based on the updated location attribute tag set.

6. The method according to claim 5, characterized in that, The step of determining the target location attribute label corresponding to the target location information based on the plurality of environmental images includes: Perform image semantic recognition on the plurality of environmental images to obtain the target environment identifiers corresponding to the plurality of environmental images; The target location attribute label corresponding to the target location is determined based on the target environment identifier.

7. An information push device, characterized in that, include: The acquisition unit is used to acquire the location information of the target terminal; The acquisition unit is used to acquire a target user profile of a target user associated with the target terminal, wherein the target user profile is obtained by fusing a static user profile and a dynamic user profile of the target user, and the dynamic user profile is constructed from the historical displacement information of the target user; wherein the historical displacement information is high-precision global positioning system trajectory data collected by the vehicle-mounted device. The process of constructing the dynamic user profile includes: Extract displacement data for the target time period from the historical displacement information; convert the displacement data for the target time period into a displacement trajectory image; Extract a target trajectory sub-image that conforms to a preset trajectory point density rule from the displacement trajectory image, wherein the preset trajectory point density rule is used to identify and extract dense trajectory point regions corresponding to the target user's dwell behavior from the displacement trajectory image, as the target trajectory sub-image; The trajectory point data in the target trajectory sub-image is classified by a preset classification decision tree to obtain the set of location attribute labels corresponding to the target trajectory sub-image. The preset classification decision tree is pre-constructed based on the point of interest classification standard of open source map data and is used to perform semantic classification of the geographical area where the trajectory points are located. The dynamic user profile is constructed based on the set of location attribute tags; The identification unit is used to identify the set of target tags corresponding to the location information; The determining unit is used to determine the information to be recommended based on the target user profile and the target tag set, and push the information to be recommended to the target terminal.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium is computer-readable and stores a plurality of instructions, which are adapted for a processor to load and execute the steps of the information push method according to any one of claims 1 to 6.

Images