Message pushing method and device, electronic equipment and computer readable storage medium

Abstract

The invention is suitable for the technical field of data processing, and provides a message pushing method and device, electronic equipment and a computer readable storage medium. The message pushing method comprises the steps of obtaining historical interaction data of a target object on a network application, terminal environment data corresponding to a target terminal where the target object is currently located, and current operation parameters corresponding to the target terminal; determining a recommendation score corresponding to each candidate message in the candidate message set by adopting an information decision model according to the historical interaction data, the terminal environment data and the current operation parameters; and based on the recommendation score, determining a target message corresponding to the target object, and pushing the target message to the target terminal. According to the method, user requirements can be accurately matched, it is ensured that message pushing adapts to the device state of the device where the user is located, invalid pushing is reduced, personalized pushing of the message can be improved, the recommendation efficiency of the 5G message reading message can be improved, and then the user satisfaction degree is improved.

Description

Technical Field

[0001] This application belongs to the field of data processing technology, and in particular relates to a message push method, apparatus, electronic device and computer-readable storage medium. Background Technology

[0002] With the rapid development of fifth-generation mobile communication (5G) technology and the widespread adoption of smartphones, 5G messaging has become a key application in the mobile communication field. 5G messaging is an upgraded service of traditional SMS, possessing the advantages of SMS such as its native nature, strong reminders, timely delivery, and high efficiency, while also enabling rich media and scene-based features, gradually becoming a new platform for internet services.

[0003] However, the relevant technologies mainly push content based on basic rules (such as user subscription information or simple classification), which lacks deep personalization and real-time performance, seriously affecting the recommendation efficiency of 5G message reading, and thus leading to low user satisfaction. Summary of the Invention

[0004] This application provides a message push method, apparatus, electronic device, and computer-readable storage medium, which can improve the personalized push of messages and the recommendation efficiency of 5G message reading, thereby improving user satisfaction.

[0005] In a first aspect, embodiments of this application provide a message push method, including:

[0006] Acquire the target object's historical interaction data on network applications, the terminal environment data corresponding to the target terminal where the target object is currently located, and the current operating parameters corresponding to the target terminal;

[0007] Based on the historical interaction data, the terminal environment data, and the current operating parameters, an information decision model is used to determine the recommendation score for each candidate message in the candidate message set.

[0008] Based on the recommendation score, the target message corresponding to the target object is determined, and the target message is pushed to the target terminal.

[0009] In some embodiments, determining the recommendation score for each candidate message in the candidate message set using an information decision model based on the historical interaction data, the terminal environment data, and the current operating parameters includes: determining the interest preference information of the target object based on the historical interaction data; determining the current scene information of the target object based on the terminal environment data; determining the terminal performance information corresponding to the target terminal based on the current operating parameters; and determining the recommendation score for each candidate message in the candidate message set using the information decision model based on the interest preference information, the current scene information, and the terminal performance information.

[0010] In some embodiments, determining the interest preference information of the target object based on the historical interaction data includes: extracting features from the historical interaction data using an interest classification model to obtain the interaction behavior features of the target object, wherein the interest classification model includes multiple preset interest categories; determining the category confidence of the target object under each preset interest category based on the interaction behavior features; and determining the interest preference information of the target object based on the category confidence.

[0011] In some embodiments, determining the current scene information of the target object based on the terminal environment data includes: extracting features from the terminal environment data using a scene classification model to obtain the terminal environment features of the target terminal, wherein the terminal environment features include the time features, location features, and device motion state features corresponding to the target terminal; classifying the scene of the target object based on the terminal environment features to obtain the current scene information.

[0012] In some embodiments, determining the terminal performance information corresponding to the target terminal based on the current operating parameters includes: extracting features from the current operating parameters using a terminal performance classification model to obtain the current performance features of the target terminal, wherein the current performance features include network bandwidth features, power consumption features, and load features; and classifying the current performance attributes of the target terminal based on the current performance features to obtain the terminal performance information.

[0013] In some embodiments, determining the recommendation score for each candidate message in the candidate message set using the information decision model based on the interest preference information, the current scene information, and the terminal performance information includes: fusing the interest preference information, the current scene information, and the terminal performance information to obtain multi-dimensional fused information; using the information decision model to determine the matching degree between each candidate message and the multi-dimensional fused information; and determining the recommendation score for each candidate message based on the matching degree.

[0014] In some embodiments, after pushing the target message to the target terminal, the method further includes: obtaining interaction feedback data of the target object in response to the target message, wherein the interaction feedback data indicates the target object's click and browsing behavior on the target message; and updating the model parameters of the information decision model based on the interaction feedback data to obtain an updated information decision model.

[0015] Secondly, embodiments of this application provide a message push device, including:

[0016] The data acquisition unit is used to acquire the historical interaction data of the target object on the network application, the terminal environment data corresponding to the target terminal where the target object is currently located, and the current operating parameters corresponding to the target terminal;

[0017] The data processing unit is used to determine the recommendation score corresponding to each candidate message in the candidate message set by using an information decision model based on the historical interaction data, the terminal environment data and the current operating parameters.

[0018] The information push unit is used to determine the target message corresponding to the target object based on the recommendation score, and push the target message to the target terminal.

[0019] Thirdly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of any of the above-described message push methods.

[0020] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by an electronic device, enables the electronic device to implement the message push method described in any of the preceding claims.

[0021] Fifthly, embodiments of this application provide a computer program product that, when run on a terminal device, causes the terminal device to execute the message push method described in any one of the first aspects.

[0022] It is understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here.

[0023] The beneficial effects of this application embodiment compared with the prior art are as follows: This application embodiment evaluates the adaptability of each candidate information in the candidate message set to the target object by comprehensively considering the historical interaction data of the target object, the terminal environment data of the target terminal where the target object is currently located, and the current operating parameters of the target terminal. This can accurately match user needs and ensure that the message push is adapted to the device status of the user's device, reducing invalid pushes. Through the above method, personalized message push can be improved, the recommendation efficiency of 5G message reading can be improved, and thus user satisfaction can be improved. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0025] Figure 1 This is a schematic diagram illustrating an application scenario of a message push method provided in an embodiment of this application;

[0026] Figure 2 This is a flowchart illustrating a message push method provided in an embodiment of this application;

[0027] Figure 3 This is a schematic diagram of a process for determining the recommendation score of candidate messages provided in an embodiment of this application;

[0028] Figure 4 This is another flowchart illustrating the process of determining the recommendation score of candidate messages provided in an embodiment of this application;

[0029] Figure 5 This is a schematic diagram of the target message determination process provided in the embodiments of this application;

[0030] Figure 6 This is a schematic diagram of the process of pushing target messages to a target terminal according to an embodiment of this application;

[0031] Figure 7 This is another flowchart illustrating the message push method provided in the embodiments of this application;

[0032] Figure 8 This is a schematic diagram illustrating the updating of the information decision-making model provided in an embodiment of this application;

[0033] Figure 9 This is a schematic diagram of the structure of the message push device provided in the embodiments of this application;

[0034] Figure 10 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0035] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0036] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0037] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0038] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0039] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0040] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0041] 5G messaging supports rich media messaging and interactive content distribution, making it a key application in mobile communications. However, existing technologies primarily rely on basic rules (such as user subscription information or simple categorization) for message push, lacking deep personalization and real-time capabilities. The main drawbacks of these existing technologies include:

[0042] (1) The user behavior analysis capability is weak, relying only on historical click data, and cannot process massive heterogeneous data (such as browsing habits and social interactions), resulting in low recommendation accuracy and serious message overload problems;

[0043] (2) The user profile is poorly constructed and does not take into account multi-dimensional scenarios (such as time, location, and device status), making it difficult to achieve personalized 5G message distribution for each user.

[0044] (3) Insufficient terminal tagging, failure to use artificial intelligence (AI) to intelligently tag device characteristics (such as network bandwidth and battery status), affecting the timeliness and adaptability of message distribution;

[0045] (4) The decision-making process lacks AI model support, relies on static algorithms, cannot dynamically optimize recommendation strategies, and is difficult to cope with complex changes in user needs, resulting in resource waste and a decline in user experience.

[0046] To address this, embodiments of this application provide a message push method, apparatus, electronic device, and computer-readable storage medium, which can improve personalized message push and increase the recommendation efficiency of 5G message reading, thereby improving user satisfaction. The message push apparatus can be integrated into an electronic device, such as a server.

[0047] The server can be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud pre-built databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, network acceleration services (Content Delivery Network, CDN), as well as big data and artificial intelligence platforms.

[0048] Figure 1 This illustration shows an application scenario diagram of a message push method provided in an embodiment of this application. For example... Figure 1 As shown, taking an example where the message push device is integrated into an electronic device, and the electronic device acts as a server, the target terminal (client) where the target object (user) is located can establish a communication connection with the server. The target terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, smart TV, in-vehicle terminal, etc., but is not limited to these. Based on this, the server can push 5G message reading messages to the target terminal where the target object is located. Specifically, the server can obtain the target object's historical interaction data on network applications, the terminal environment data corresponding to the target terminal where the target object is currently located, and the current operating parameters of the target terminal; based on the historical interaction data, terminal environment data, and current operating parameters, an information decision model is used to determine the recommendation score corresponding to each candidate message in the candidate message set; based on the recommendation score, the target message corresponding to the target object is determined, and the target message is pushed to the target terminal.

[0049] It is understood that, in the specific implementation of this application, the target object's historical interaction data on network applications, the terminal environment data corresponding to the target terminal where the target object is currently located, the current operating parameters of the target terminal, and other related data are involved. When the following embodiments of this application are applied to specific products or technologies, permission or consent is required, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0050] The following sections provide detailed descriptions of each example. It should be noted that the order in which the embodiments are described is not intended to limit the preferred order of the embodiments.

[0051] Figure 2 A flowchart illustrating a message push method provided in an embodiment of this application is shown. Figure 2 As shown, the specific process of this message push method is as follows:

[0052] S101. Obtain the target object's historical interaction data on network applications, the terminal environment data corresponding to the target terminal where the target object is currently located, and the current running parameters corresponding to the target terminal.

[0053] Historical interaction data can be understood as user log data generated by the target object during the use of various types of online applications (such as social media applications, online shopping applications, etc.). For example, it can be user log data corresponding to the target object browsing or purchasing goods in online shopping applications, or it can be structured and unstructured user log data such as the target object's browsing history, browsing duration, interaction frequency, and click records of media content in social media applications.

[0054] Historical interaction data can be valid user log data generated by the target object before the current moment. In this embodiment, all user log data generated by the target object before the current moment can be collected in real time through an application programming interface (API). Considering that there may be a lot of invalid information in the user log data, the collected user log data can be cleaned to filter noise (such as invalid clicks, invalid browsing, etc.) to obtain valid user log data (historical interaction data of the target object on the network application).

[0055] The terminal environment data can be understood as environmental data reflecting the current environment of the target terminal. Since the target terminal is the terminal device currently in which the target object is located, the terminal environment data can also reflect the current environment of the target object. Terminal environment data can include the target terminal's current time, location (e.g., multi-dimensional geographic information data generated based on location services, i.e., LBS data), device motion status, etc. Device motion status can include running, walking normally, exercising, riding in transportation, etc. In this embodiment, sensor data (such as accelerometer, GPS, etc.) corresponding to the target terminal in real time can be collected, and the collected sensor data can be cleaned to obtain the terminal environment data.

[0056] The current operating parameters can be understood as parameters reflecting the device operating status of the target terminal at the current moment. These parameters may include network bandwidth, battery level, CPU load, and other corresponding parameter information. In this embodiment, the operating parameters of the target terminal can be detected in real time, and the current operating parameters of the target terminal can be obtained by calling a system-level API.

[0057] S102. Based on historical interaction data, terminal environment data, and current operating parameters, an information decision model is used to determine the recommendation score corresponding to each candidate message in the candidate message set.

[0058] In this embodiment, historical interaction data, terminal environment data, and current operating parameters can be combined to calculate and evaluate the recommendation score corresponding to each candidate message in the candidate message set, thereby determining the target message to be pushed to the target terminal based on the recommendation score corresponding to each candidate message.

[0059] The candidate message set is a collection of 5G message messages to be pushed, and each candidate message in the candidate message set is a 5G message to be pushed.

[0060] Figure 3 This illustration shows a flowchart of a process for determining the recommendation score of candidate messages according to an embodiment of this application. Figure 3 As shown, step S102, based on historical interaction data, terminal environment data, and current operating parameters, uses an information decision model to determine the recommendation score corresponding to each candidate message in the candidate message set, which may include:

[0061] S1021. Determine the target object's interest and preference information based on historical interaction data.

[0062] Figure 4 This illustration shows another flowchart illustrating the process of determining the recommendation score of candidate messages according to an embodiment of this application. Figure 4As shown, the interest and preference information of a target object can be determined based on historical interaction data. Specifically, determining the interest and preference information of a target object based on historical interaction data can include: extracting features from historical interaction data using an interest classification model to obtain the interaction behavior features of the target object; the interest classification model includes multiple preset interest categories; determining the category confidence of the target object under each preset interest category based on the interaction behavior features; and determining the interest and preference information of the target object based on the category confidence.

[0063] In this embodiment, an interest classification model can be used to process the historical interaction data of the target object, so as to filter the target user's target interest category from multiple preset interest categories, thereby obtaining the target object's interest preference information. The preset interest categories can be set according to actual conditions, for example, they can be technology enthusiasts, fashion consumers, high-value users, potential churned users, etc.

[0064] Interest classification models can extract interactive behavior features of target objects from historical interaction data and calculate the category confidence of the target object in each preset interest category using clustering algorithms (such as K-means clustering). Category confidence can be understood as the score of the target object belonging to each preset interest category; the higher the score, the higher the probability that the target object belongs to that preset interest category.

[0065] After obtaining the category confidence score of the target object under each preset interest category, the interest preference information of the target object can be determined based on the category confidence score, that is, which preset interest category the target object belongs to.

[0066] S1022. Determine the current scene information of the target object based on the terminal environment data.

[0067] like Figure 4 As shown, the current scene information of the target object can be determined based on terminal environment data. Specifically, determining the current scene information of the target object based on terminal environment data can include: using a scene classification model to extract features from the terminal environment data to obtain the terminal environment features of the target terminal, which include the current time features, location features, and device motion state features of the target terminal; and classifying the scene in which the target object is currently located based on the terminal environment features to obtain the current scene information.

[0068] In this embodiment, a scene classification model can be used to process terminal environment data to identify the current scene of the target object, thereby determining the current scene information corresponding to the target object. Terminal environment features may include the current time features, location features, and device motion state features of the target terminal.

[0069] The scene classification model can extract the terminal environment features of the target terminal from the terminal environment data, and identify the terminal environment features through random forest algorithm or other classification algorithms (such as BERT-based natural language processing algorithms), and filter the current scene information of the target object from the preset scene dataset. The preset scene dataset can include multiple preset scenes, such as commuting scenes, leisure at home scenes, work scenes, entertainment scenes, travel scenes, etc.

[0070] Scene classification models can predict the probability value of a target terminal belonging to each preset scene based on extracted terminal environment features. Then, based on the scene probability value for each preset scene, the model determines the preset scene that the target terminal is most likely to match, thus obtaining the current scene information. The higher the scene probability value corresponding to a preset scene, the more likely the target terminal (target object) is to belong to that preset scene. For example, if the scene classification model, after processing data based on terminal environment features, determines that the probability value for a commuting scene is 80%, the probability value for an entertainment scene is 10%, and the probability value for a work scene is 10%, then it can be determined that the target terminal (target object) is most likely currently in a commuting scene.

[0071] S1023. Determine the terminal performance information corresponding to the target terminal based on the current operating parameters.

[0072] like Figure 4 As shown, the terminal performance information corresponding to the target terminal can be determined based on the current operating parameters. Specifically, determining the terminal performance information corresponding to the target terminal based on the current operating parameters can include: using a terminal performance classification model to extract features from the current operating parameters to obtain the current performance characteristics of the target terminal, which include network bandwidth characteristics, power consumption characteristics, and load characteristics; and classifying the current performance attributes of the target terminal based on the current performance characteristics to obtain the terminal performance information.

[0073] In this embodiment, a terminal performance classification model can be used to process the current operating parameters in order to classify the current performance attributes of the target terminal and obtain terminal performance information.

[0074] The terminal performance classification model can extract the current performance characteristics of the target terminal from the current operating parameters. These current performance characteristics can include network bandwidth, battery level, and load characteristics. Based on these extracted characteristics, the model can use algorithms such as decision trees to determine the terminal's performance information. The feature values ​​corresponding to each performance characteristic reflect the target terminal's operating status at the current moment. For example, based on different current performance characteristics, the target terminal's operating status can be divided into a high-bandwidth state and a low-battery state. In the high-bandwidth state, the target terminal can be called a high-performance terminal, while in the low-battery state, it can be called a restricted terminal.

[0075] For example, when the network bandwidth is greater than 50 Mbps and the battery level is greater than 50%, the terminal performance information can be determined to be a high-performance terminal; conversely, when the network bandwidth is less than or equal to 50 Mbps or the battery level is less than or equal to 50%, the terminal performance information can be determined to be a limited terminal.

[0076] S1024. Based on interest preference information, current scene information, and terminal performance information, an information decision model is used to determine the recommendation score corresponding to each candidate message in the candidate message set.

[0077] like Figure 4 As shown, an information decision model can be used to comprehensively process interest preference information, current scene information, and terminal performance information to calculate the recommendation score for each candidate message in the candidate message set. The recommendation score can be understood as the probability of the target object clicking on the candidate message as predicted by the information decision model. A higher recommendation score indicates a higher probability of the target object clicking on the candidate message; conversely, a lower recommendation score indicates a lower probability of the target object clicking on the candidate message.

[0078] In this embodiment of the application, the recommendation score corresponding to each candidate message in the candidate message set is determined by using an information decision model based on interest preference information, current scene information, and terminal performance information. This may include: fusing interest preference information, current scene information, and terminal performance information to obtain multi-dimensional fused information; using an information decision model to determine the matching degree between each candidate message and the multi-dimensional fused information; and determining the recommendation score corresponding to each candidate message based on the matching degree.

[0079] Figure 5 A schematic diagram illustrating the target message determination process provided in an embodiment of this application is shown. Figure 5 As shown, the information decision-making model can be based on a reinforcement learning framework (such as Q-learning) to fuse interest preference information, current scene information, and terminal performance information to obtain multi-dimensional fused information; and based on the multi-dimensional fused information, the recommendation score corresponding to each candidate message can be determined.

[0080] S103. Based on the recommendation score, determine the target message corresponding to the target object and push the target message to the target terminal.

[0081] After obtaining the recommendation score for each candidate message, the target message corresponding to the target object can be selected from the candidate message set based on the recommendation score of each candidate message. Specifically, the recommendation scores of each candidate message can be sorted, and one or more candidate messages with the highest recommendation scores can be selected based on the sorting results. The selected candidate messages are the target messages.

[0082] In this context, target messages can be understood as personalized 5G messaging content selected for a specific target audience. These target messages can be various personalized text and image messages adapted to the target audience. For example, if interest preference information indicates that the target audience is a technology enthusiast, current scene information indicates that the target audience is currently commuting, and terminal performance information indicates that the target terminal is currently in a high-bandwidth state, then a short technology news summary from the candidate message set can be used as the target information.

[0083] For example, when interest preference information shows that the target is a fashion consumer, current scene information shows that the target is currently in a leisurely home setting, and terminal performance information shows that the target terminal is currently in a high-performance state, promotional information in the candidate message set can be used as the target information.

[0084] In this embodiment of the application, for the selected target message, a suitable message template can be selected from the message template library according to the message content of the target message, and the target message can be packaged according to the template form of the message template, so that the same target message content can be packaged into various 5G message content.

[0085] Figure 6 This illustration shows a flowchart of pushing target messages to a target terminal according to an embodiment of this application. Figure 6 As shown, for the selected target messages, the messages can be distributed to the target terminal where the target object is located through the 5G core network. In this process, edge node computing and Quality of Service (QoS) detection can be performed first, and then the messages can be distributed to the target terminal where the target object is located through the 5G core network.

[0086] Figure 7 This illustration shows yet another flowchart of the message push method provided in an embodiment of this application. Figure 7 As shown, after pushing the target message to the target terminal, the process may further include: obtaining the target object's interaction feedback data in response to the target message, the interaction feedback data indicating the target object's click and browsing behavior regarding the target message; and updating the model parameters of the information decision model based on the interaction feedback data to obtain the updated information decision model.

[0087] Interactive feedback data can be understood as how the target object processes the target message after seeing it pushed to its target terminal. For example, the target object can click to enter the target message, or it can ignore the target message, and so on.

[0088] Figure 8 This illustration shows a schematic diagram of updating an information decision-making model according to an embodiment of this application. For example... Figure 8 As shown, the interactive behavior of the target object can be detected in real time. After obtaining interaction feedback data (such as clicks or ignores), different push strategies can be compared based on the A / B testing component to ensure continuous algorithm improvement. This process can adopt a microservice architecture to ensure high concurrency processing capabilities, such as handling more than 100,000 requests per second. Furthermore, the overall implementation of this process can be deployed using Docker containers, supporting elastic scaling.

[0089] The message push method provided in this application makes decisions on target information by comprehensively considering the target object's historical interaction data, the target object's current terminal environment data, and the target terminal's current operating parameters. The historical interaction data can be used to construct user profiles and identify the target object's interests and preferences; the terminal environment data can accurately locate the target object's current scene; and the current operating parameters can determine the target terminal's current device operating status. This achieves the following technical effects:

[0090] (1) Improve the accuracy of message recommendation: The target message determined by the above method can accurately match user needs, and the click rate of users for the target message can be increased to 15% (about 3 times higher than the existing technology).

[0091] (2) Resource optimization effect: It can ensure that message push is adapted to the device status of the user's device, reduce invalid pushes, increase network resource utilization by 30%, and reduce battery consumption by 20%.

[0092] (3) Enhanced user experience: The decision-making AI algorithm of the information decision model can support dynamic personalization, increasing user satisfaction by 25% and reducing complaints about message overload by 40%.

[0093] (4) Realization of commercial value: The personalized distribution mechanism can improve the advertising conversion rate and increase ROI (return on investment) by 50%, which is suitable for large-scale commercial applications.

[0094] The message push method provided in this application can be applied to the following fields:

[0095] In the field of mobile communication services: Telecommunication operators can use the message push method provided in the embodiments of this application on the 5G message reading platform to realize personalized message push (such as customized package recommendations) and improve user stickiness.

[0096] In the field of content distribution and advertising: content platforms can utilize the message push method provided in the embodiments of this application for intelligent news or video recommendations, and distribute rich media 5G messaging messages in conjunction with user profiles to improve user engagement and advertising revenue. The application can be embedded into a mobile application using a software development kit.

[0097] In the e-commerce sector: E-commerce platforms can use the message push method provided in this application to analyze user shopping behavior, distribute personalized promotional 5G messages (such as limited-time offers), and drive conversion rates. Application methods may include integration with a customer relationship management system to trigger contextualized messages in real time.

[0098] In the education and public service sectors: online education institutions or public service platforms can utilize the message push method provided in this application's embodiments to recommend learning resources or public service information, and distribute interactive 5G message reading based on scenario analysis (such as students' home status). Potential applications can be extended to smart city projects.

[0099] Corresponding to the message push method described in the above embodiments, this application also provides a message push device. Figure 9 A schematic diagram of the message push device provided in an embodiment of this application is shown. Figure 9 As shown, the message push device may include a data acquisition unit 201, a data processing unit 202, and an information push unit 203, as follows:

[0100] (1) Data acquisition unit 201;

[0101] The data acquisition unit 201 is used to acquire the historical interaction data of the target object on the network application, the terminal environment data corresponding to the target terminal where the target object is currently located, and the current running parameters corresponding to the target terminal.

[0102] (2) Data processing unit 202;

[0103] The data processing unit 202 is used to determine the recommendation score corresponding to each candidate message in the candidate message set based on historical interaction data, terminal environment data and current operating parameters using an information decision model.

[0104] For example, the data processing unit 202 can be used to determine the interest and preference information of the target object based on historical interaction data; determine the current scene information of the target object based on terminal environment data; determine the terminal performance information corresponding to the target terminal based on current operating parameters; and determine the recommendation score corresponding to each candidate message in the candidate message set based on the interest and preference information, current scene information and terminal performance information using an information decision model.

[0105] For example, data processing unit 202 can be used to extract features from historical interaction data using an interest classification model to obtain the interaction behavior features of the target object. The interest classification model includes multiple preset interest categories. Based on the interaction behavior features, the category confidence of the target object under each preset interest category is determined. Based on the category confidence, the interest preference information of the target object is determined.

[0106] For example, the data processing unit 202 can be used to extract features from the terminal environment data using a scene classification model to obtain the terminal environment features of the target terminal. The terminal environment features include the current time features, location features, and device motion state features of the target terminal. Based on the terminal environment features, the current scene of the target object is classified to obtain the current scene information.

[0107] For example, the data processing unit 202 can be used to extract features from the current operating parameters using a terminal performance classification model to obtain the current performance features of the target terminal, including network bandwidth features, power consumption features, and load features; and to classify the current performance attributes of the target terminal based on the current performance features to obtain terminal performance information.

[0108] For example, the data processing unit 202 can be used to fuse interest preference information, current scene information and terminal performance information to obtain multi-dimensional fused information; use an information decision model to determine the matching degree between each candidate message and the multi-dimensional fused information; and determine the recommendation score corresponding to each candidate message based on the matching degree.

[0109] (3) Information push unit 203;

[0110] The information push unit 203 is used to determine the target message corresponding to the target object based on the recommendation score, and push the target message to the target terminal.

[0111] In some embodiments, the message push device may further include a feedback update unit 204.

[0112] The feedback update unit 204 is used to acquire the interaction feedback data of the target object in response to the target message. The interaction feedback data indicates the target object's click and browsing status of the target message. Based on the interaction feedback data, the model parameters of the information decision model are updated to obtain the updated information decision model.

[0113] In practice, each of the above units can be implemented as an independent entity or can be arbitrarily combined to be implemented as the same or several entities. For the specific implementation of each of the above units, please refer to the previous method embodiments, which will not be repeated here.

[0114] It should be noted that the information interaction and execution process between the above units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0115] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units is used as an example. In practical applications, the above functions can be assigned to different functional units as needed, that is, the internal structure of the device can be divided into different functional units to complete all or part of the functions described above. The functional units in the embodiments can be integrated into one processing unit, or they can exist as separate physical units, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units in the above-described device can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0116] Figure 10 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 10 As shown, the electronic device 3 of this embodiment includes: at least one processor 30 ( Figure 10 (Only one is shown in the image), memory 31, and computer program 32 stored in memory 31 and executable on at least one processor 30. When processor 30 executes computer program 32, it implements the steps in any of the above-described message push method embodiments. Alternatively, when processor 30 executes computer program 32, it implements the steps in any of the above-described message push method embodiments.

[0117] The electronic device 3 may include, but is not limited to, a processor 30 and a memory 31. Those skilled in the art will understand that... Figure 10This is merely an example of electronic device 3 and does not constitute a limitation on electronic device 3. It may include more or fewer components than shown in the figure, or combine certain components, or different components. For example, it may also include input / output devices, network access devices, etc.

[0118] The processor 30 can be a central processing unit (CPU), but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0119] In some embodiments, memory 31 may be an internal storage unit of electronic device 3, such as a hard disk or memory of electronic device 3. In other embodiments, memory 31 may be an external storage device of electronic device 3, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on electronic device 3. Furthermore, memory 31 may include both internal storage units and external storage devices of electronic device 3. Memory 31 is used to store operating system, application programs, bootloader, data, and other programs, such as program code of computer programs. Memory 31 may also be used to temporarily store data that has been output or will be output.

[0120] This application also provides a computer-readable storage medium storing a computer program. When the computer program is executed by an electronic device, the electronic device performs the steps described in the above-described message push method embodiments.

[0121] This application provides a computer program product, which includes a computer program. When the computer program is executed by an electronic device, it enables the electronic device to implement the steps in the above-described message push method embodiments.

[0122] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable storage medium can include at least: any entity or device capable of carrying computer program code to a device / electronic device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable storage media cannot be electrical carrier signals or telecommunication signals.

[0123] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0124] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0125] In the embodiments provided in this application, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0126] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0127] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A message pushing method, characterized by, The method comprises the following steps: obtaining historical interaction data of a target object on a network application, terminal environment data corresponding to a target terminal where the target object is currently located, and current running parameters corresponding to the target terminal; determining a recommendation score corresponding to each candidate message in a candidate message set by using an information decision model according to the historical interaction data, the terminal environment data and the current running parameters; determining a target message corresponding to the target object based on the recommendation score, and pushing the target message to the target terminal.

2. The message pushing method of claim 1, wherein, The method comprises the following steps: determining interest preference information of the target object according to the historical interaction data; determining current scene information where the target object is located according to the terminal environment data; determining terminal performance information corresponding to the target terminal according to the current running parameters; determining the recommendation score corresponding to each candidate message in the candidate message set by using the information decision model based on the interest preference information, the current scene information and the terminal performance information.

3. The message pushing method of claim 2, wherein, The method comprises the following steps: extracting features of the historical interaction data by using an interest classification model to obtain interaction behavior features of the target object, the interest classification model comprising a plurality of preset interest categories; determining a category confidence of the target object in each preset interest category according to the interaction behavior features; determining the interest preference information of the target object based on the category confidence.

4. The message pushing method of claim 2, wherein, The method comprises the following steps: extracting features of the terminal environment data by using a scene classification model to obtain terminal environment features of the target terminal, the terminal environment features comprising time features, location features and device motion state features corresponding to the target terminal at present; classifying a scene where the target object is currently located according to the terminal environment features to obtain the current scene information.

5. The message pushing method of claim 2, wherein, The method comprises the following steps: extracting features of the current running parameters by using a terminal performance classification model to obtain current performance features of the target terminal, the current performance features comprising network bandwidth features, power features and load features; classifying performance attributes of the target terminal at present according to the current performance features to obtain the terminal performance information. 6.The message pushing method of claim 2, wherein, The method comprises the following steps: fusing the interest preference information, the current scene information and the terminal performance information to obtain multi-dimensional fusion information; determining a matching degree between each candidate message and the multi-dimensional fusion information by using the information decision model; According to the matching degree, a recommendation score corresponding to each candidate message is determined.

7. The message pushing method of claim 1, wherein, After the target message is pushed to the target terminal, the method further includes: Obtaining interaction feedback data of the target object for the target message, the interaction feedback data indicating a click browsing condition of the target object for the target message; Based on the interaction feedback data, updating model parameters of the information decision model to obtain an updated information decision model.

8. A message pushing apparatus characterized by comprising: The method comprises: a data collection unit configured to obtain historical interaction data of a target object on a network application, terminal environment data corresponding to a target terminal in which the target object is currently located, and current running parameters corresponding to the target terminal; a data processing unit configured to determine, according to the historical interaction data, the terminal environment data, and the current running parameters, a recommendation score corresponding to each candidate message in a candidate message set by using an information decision model; an information pushing unit configured to determine a target message corresponding to the target object based on the recommendation score, and push the target message to the target terminal.

9. An electronic device, comprising: The computer program is executed by the processor to implement the message pushing method according to any one of claims 1 to 7. The computer program is executed by the processor to implement the message pushing method according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 9. ​

Images