Message pushing method, computing device, storage medium and computer program product

Abstract

The embodiment of the invention provides a message pushing method, computing equipment, a computer readable storage medium and a computer program product. The message push method comprises the following steps: acquiring at least one push message requested to be sent to a client by a target service account; determining a preference matching degree of the client to the target service account according to the at least one push message by using a message identification model; under the condition that the preference matching degree meets a filtering condition, generating a message filtering instruction containing target identification information of the target service account; and sending the message filtering instruction to the client in a silent pushing mode. According to the technical scheme provided by the embodiment of the invention, the instruction server actively utilizes the message identification model to accurately predict the interest degree of the client in the service account, so that the pushed message is accurately shielded, and frequent interaction between the client and the server is reduced.

Description

Technical Field

[0001] Embodiments of the present application relate to the field of artificial intelligence technology, and in particular, to a message push method, a computing device, a computer-readable storage medium, and a computer program product. Background Art

[0002] With the popularization of the mobile Internet and intelligent devices, service accounts maintain interaction with users by pushing messages. A service account may refer to a virtual account created and operated by an enterprise, a platform, or a developer, and is used to provide specific services or information to a client. For example, in the e-commerce field, a service account may be an official account of a certain brand; in the social field, it may be a certain official account or subscription account. Service accounts usually use a message server to send push messages to the client to deliver the latest information, promotional activities, system notifications, and other content to the client.

[0003] However, with the increase in push messages, the problem of message overload faced by users has become increasingly serious. In particular, a large number of irrelevant or low-value operation messages interfere with users and affect the user experience. Summary of the Invention

[0004] Embodiments of the present application provide a message push method, a computing device, a computer-readable storage medium, and a computer program product.

[0005] In a first aspect, a message push method provided in an embodiment of the present application is applied to an instruction server, and the method includes:

[0006] Obtain at least one push message that a target service account requests to send to a client;

[0007] Use a message recognition model to determine a preference matching degree of the client for the target service account according to the at least one push message, where the message recognition model is trained with sample push data and sample behavior data of the client for the sample push data as input data, and the sample preference matching degree of the client for the sample service account corresponding to the sample push data as a training label;

[0008] Generate a message filtering instruction including target identification information of the target service account when the preference matching degree meets a filtering condition;

[0009] Send the message filtering instruction to the client in a silent push manner, so that when the target client receives the at least one push message, it determines the service account corresponding to the at least one push message, and when the service account is the target service account, performs filtering processing on the at least one push message.

[0010] Second aspect, an embodiment of the present application provides a message pushing method, which is applied to a client, and the method includes:

[0011] Receiving a first message filtering instruction sent by a server in a silent push manner, where the message filtering instruction contains target identification information of a target service account. Among them, the preference matching degree of the client for the target service account is lower than a preset threshold, and the preference matching degree is determined by a message recognition model. The message recognition model is trained with sample push data and sample behavior data of the client for the sample push data as input data, and the sample preference matching degree of the client for the sample service account corresponding to the sample push data as a training label;

[0012] Receiving a push message to be processed;

[0013] Determining first identification information of a service account corresponding to the push message to be processed;

[0014] Determining whether the target identification information matches the first identification information;

[0015] If so, filtering the push message to be processed.

[0016] Third aspect, an embodiment of the present application provides a message pushing device, which is applied to an instruction server, and the device includes:

[0017] A message acquisition module, configured to acquire at least one push message requested to be sent by a target service account to a client;

[0018] A preference determination module, configured to use a message recognition model to determine the preference matching degree of the client for the target service account according to the at least one push message. Among them, the message recognition model is trained with sample push data and sample behavior data of the client for the sample push data as input data, and the sample preference matching degree of the client for the sample service account corresponding to the sample push data as a training label;

[0019] An instruction generation module, configured to generate a message filtering instruction containing the target identification information of the target service account when the preference matching degree meets a filtering condition;

[0020] An instruction push module, configured to send the message filtering instruction to the client in a silent push manner, so that the target client determines a service account corresponding to the at least one push message when receiving the at least one push message, and filters the at least one push message when the service account is the target service account.

[0021] Fourth aspect, an embodiment of the present application provides a message push device, which is applied to a client. The device includes:

[0022] An instruction receiving module, configured to receive a first message filtering instruction sent by a server in a silent push manner. The message filtering instruction includes target identification information of a target service account. Among them, the preference matching degree of the client for the target service account is lower than a preset threshold. The preference matching degree is determined by a message recognition model. The message recognition model is trained with sample push data and sample behavior data of the client for the sample push data as input data, and the sample preference matching degree of the client for the sample service account corresponding to the sample push data as a training label.

[0023] A message receiving module, configured to receive a to-be-processed push message.

[0024] An identification information determination module, configured to determine first identification information of the service account corresponding to the to-be-processed push message.

[0025] A matching module, configured to determine whether the target identification information matches the first identification information.

[0026] A filtering module, configured to, if so, perform filtering processing on the to-be-processed push message.

[0027] Fifth aspect, an embodiment of the present application provides a computing device, including a processing component and a storage component;

[0028] The storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the message push method provided by the embodiment of the present application.

[0029] Sixth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processing component, the message push method provided by the embodiment of the present application is implemented.

[0030] Seventh aspect, an embodiment of the present application provides a computer program product, including a computer program / instructions. When the computer program / instructions are executed by a processing component, the message push method provided by the embodiment of the present application is implemented.

[0031] In an embodiment of the present application, by adopting: obtaining at least one push message sent by a target service account to a client; using a message recognition model to determine the preference matching degree of the client for the target service account according to the at least one push message, wherein the message recognition model is trained with sample push data and sample behavior data of the client for the sample push data as input data, and using the sample preference matching degree of the client for the sample service account corresponding to the sample push data as a training label; generating a message filtering instruction including the target identification information of the target service account when the preference matching degree meets a filtering condition; and sending the message filtering instruction to the client in a silent push manner, so that when the target client receives the at least one push message, it determines the service account corresponding to the at least one push message, and when the service account is the target service account, performs filtering processing on the at least one push message. The technical solution enables the instruction server to actively and accurately predict the degree of interest of the client in the service account by using the message recognition model, thereby realizing precise shielding of push messages, reducing frequent interactions between the client and the server, and reducing network load and resource consumption. In addition, by means of the silent push technology, the shielding instruction can be sent to the background of the client without manual operation by the user, significantly improving the response speed.

[0032] These aspects or other aspects of the present application will be more clearly understood in the following description of the embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the embodiments of the present application or in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0034] Figure 1 FIG. shows a system architecture diagram to which a technical solution of an embodiment of the present application can be applied;

[0035] Figure 2 FIG. shows a flowchart of a message push method provided by an embodiment of the present application;

[0036] Figure 3 FIG. shows a flowchart of a message push method provided by another embodiment of the present application;

[0037] Figure 4 FIG. shows a block diagram of a message push device provided by an embodiment of the present application;

[0038] Figure 5The block diagram of a message push device provided by another embodiment of the present application is shown;

[0039] Figure 6 The block diagram of a computing device provided by an embodiment of the present application is shown. Detailed implementation manners

[0040] In order to enable those skilled in the art to better understand the solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application.

[0041] In some processes described in the specification, claims and above-mentioned drawings of the present application, a plurality of operations appear in a specific order. However, it should be clearly understood that these operations may not be executed in the order in which they appear herein or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions such as "first" and "second" in this article are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, nor do they limit that "first" and "second" are of different types.

[0042] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties. And the collection, use and processing of relevant data need to comply with the relevant laws, regulations and standards of relevant countries and regions, and corresponding operation entrances are provided for users to choose to authorize or refuse.

[0043] It should be noted that the technical solutions of the embodiments of the present application are applicable to a network virtual environment. The users described generally refer to "virtual users". Real users can register user accounts on the server through the registration method to obtain user identities in the network environment. In the embodiments of the present application, the same user account can be logged in to the server through different types of clients, so that the server can identify the same user. Of course, different user accounts can also be logged in to the server through different types of clients. The server stores the binding relationships of different user accounts, so that different user accounts with binding relationships can be considered as the same user.

[0044] With the popularization of mobile Internet and intelligent devices, service accounts maintain interaction with users by pushing messages. A service account may refer to a virtual account created and operated by an enterprise, platform or developer to provide specific services or information to the client. For example, in the e-commerce field, a service account may be the official account of a certain brand; in the social field, it may be a certain official account or subscription account. Service accounts usually use the message server to send push messages to the client to deliver the latest information, promotional activities, system notifications and other content to the client.

[0045] However, with the increase in push messages, the problem of message overload faced by users has become increasingly serious. In particular, a large number of irrelevant or low-value operation messages interfere with users and affect the user experience.

[0046] In related technologies, when the client does not want to receive push messages from a certain service account, it mainly relies on a push message filtering mechanism based on client-server interaction. The specific process is as follows: The user needs to manually operate the client (such as clicking the "block" button) to send a network request to the server, clearly informing the server of the service account to be blocked. This request is usually transmitted to the server through HTTP or other network protocols. After receiving the client's request, the server will generate a blocking instruction according to the request content and send the identification information of the service account to the client. After receiving the blocking instruction, the client will store it in the local database. When receiving new push messages subsequently, the client will check one by one whether the source identification information of the message is consistent with the identification information in the blocking list. If they are consistent, the message will be directly intercepted and not shown to the user.

[0047] The inventors found in the process of implementing the concept of this application that the message filtering mechanism in related technologies needs to rely on the network request and response process between the client and the server. This mechanism will cause obvious delays in case of poor network conditions, affecting the user experience. In addition, frequent network interactions and manual operations also increase the complexity of push message blocking and the operation burden of users.

[0048] To solve the technical problems existing in the related art, an embodiment of the present application provides a message pushing method, which includes: obtaining at least one push message sent by a target service account to a client; using a message recognition model to determine the preference matching degree of the client for the target service account according to the at least one push message, where the message recognition model is trained with sample push data and sample behavior data of the client for the sample push data as input data, and the sample preference matching degree of the client for the sample service account corresponding to the sample push data as the training label; generating a message filtering instruction including the target identification information of the target service account when the preference matching degree meets the filtering condition; and sending the message filtering instruction to the client in a silent push manner, so that when the target client receives the at least one push message, it determines the service account corresponding to the at least one push message, and when the service account is the target service account, performs filtering processing on the at least one push message. The technical solution enables the instruction server to actively and accurately predict the client's interest level in the service account by using the message recognition model, thereby achieving precise shielding of push messages, reducing frequent interactions between the client and the server, and reducing network load and resource consumption. In addition, with the help of the silent push technology, the shielding instruction can be sent to the client background without manual operation by the user, significantly improving the response speed.

[0049] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0050] Figure 1 FIG. shows a system architecture diagram to which a technical solution of an embodiment of the present application can be applied. The system architecture may include a client 101 and a server 102.

[0051] Among them, a connection is established between the client 101 and the server 102 through a network. The network provides a medium for the communication link between the client 101 and the server 102. The network may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc. Optionally, the server can communicate with the client through a mobile network. Optionally, the client can also establish a communication connection with the server through Bluetooth, WiFi, infrared, etc.

[0052] The client 101 can interact with the server 102 through the network to receive or send messages, etc.

[0053] Among them, the client 101 can be a browser, an APP (Application), or a web application such as an H5 (HyperText Markup Language 5) application, or a light application (also known as a mini-program, a lightweight application), or a cloud application, etc. The client 101 can be deployed in an electronic device and needs to rely on the device or certain apps in the device to run, etc. The electronic device can, for example, have a display screen and support information browsing, etc., such as a personal mobile terminal such as a mobile phone, a tablet computer, a personal computer, a desktop computer, a smart speaker, a smart watch, etc. For the sake of easy understanding, Figure 1 in Figure 1 , the client is mainly represented by the image of the device. Various other types of applications can usually be configured in the electronic device, such as human-computer dialogue applications, model training applications, text processing applications, web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc. The electronic device can refer to a device used by a user and having functions such as computing, Internet access, and communication required by the user. For example, it can be a mobile phone, a tablet computer, a personal computer, a wearable device, etc. The electronic device usually can include at least one processing component and at least one storage component. The electronic device may also include basic configurations such as a network card chip, an IO bus, and audio-video components, which are not limited in this application. Optionally, according to the implementation form of the electronic device, some peripheral devices may also be included, such as a keyboard, a mouse, a stylus, a printer, etc., which are not limited in this application.

[0054] The server 102 can include servers that provide various services. For example, a server for background training that provides support for the models used on the client 101, or a server that processes the interaction information sent by the client, etc.

[0055] It should be noted that the server 102 can be implemented as a distributed server cluster composed of multiple servers, or can be implemented as a single server. The server can also be a server of a distributed system, or a server combined with a blockchain. The server can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, Content Delivery Network (CDN), as well as big data and artificial intelligence platforms, or an intelligent cloud computing server or an intelligent cloud host with artificial intelligence technology.

[0056] It should be understood that Figure 1The numbers of the client and the server in it are only illustrative. According to the implementation requirements, there can be any number of clients and servers. For example, in the embodiments of the present application, the server may include an instruction server and a message server. Among them, the instruction server serves as a control center and is responsible for managing and coordinating operations related to message pushing. It can determine which messages should be sent to the client and which messages need to be blocked according to the preferences and behavior data of the client. The message server is specifically responsible for message distribution. A push request is initiated by a service account (such as an operation account of a certain brand), and the message is sent to the client through the message server.

[0057] The implementation details of the technical solutions in the embodiments of the present application are elaborated in detail below.

[0058] Figure 2 The flowchart of a message pushing method provided by an embodiment of the present application is shown. This method can be applied to an instruction server, such as Figure 2 As shown, this method may specifically include the following steps:

[0059] 201: Obtain at least one push message that the target service account requests to send to the client.

[0060] Among them, the target service account may include the service account that needs to be analyzed currently. For example, in an e-commerce application, a certain clothing brand may be a target service account; in a social platform, a certain official account or subscription account can also be regarded as a target service account. Each service account can have its unique identification information, which is used to distinguish different message sources.

[0061] When the target service account hopes to send a push message to the user, it can initiate a push request through the message server. The push request may include at least one push message to be sent. Each push message may include, for example, message content, message type, message characteristics, target users, etc.

[0062] The instruction server can receive the push request of the target service account from the message server and extract the push message therein.

[0063] In an implementation manner of the present application, the instruction server can obtain at least one push message that the message server requests to send to the client by means of buried point interception.

[0064] In some embodiments, obtaining at least one push message that the target service account requests to send to the client can be specifically implemented as:

[0065] Intercept at least one push message that the message server corresponding to the target service account sends to the client.

[0066] The instruction server can pre-insert monitoring code or logic in the message server to capture and record relevant events or data. In the embodiments of the present application, the data point embedding can be set on the communication link between the message server and the client, such as the node of message distribution. For example, when the message server is about to send a push message to the client, the data point embedding will be automatically triggered to capture the push request of the target service account.

[0067] The intercepted push message can be sent to the instruction server for further analysis. In the embodiments of the present application, the instruction server can combine the user's preference data and behavior model to determine whether to block the push message.

[0068] Through the method of intercepting with data point embedding, the instruction server can efficiently obtain the push messages sent from the target service account to the client, which not only realizes the real-time capture of push messages, but also provides a data basis for subsequent personalized blocking.

[0069] 202: Using a message recognition model, determine the preference matching degree of the client for the target service account according to at least one push message, wherein the message recognition model is trained with sample push data and sample behavior data of the client for the sample push data as input data, and the sample preference matching degree of the client for the sample service account corresponding to the sample push data as the training label.

[0070] In the embodiments of the present application, the message recognition model can be implemented as an algorithm based on machine learning, which can evaluate the degree of interest of the client in a certain service account according to the user's behavior data and message characteristics.

[0071] Among them, the message recognition model can be trained and generated using sample push data and client behavior data. The sample push data can include, for example, historical push messages sent by the target service account or other similar service accounts to the client in the past period of time. These historical push messages contain rich information, such as message content (text, pictures, etc.), message type (such as advertisements, notifications, etc.), message source (service account identification information), and key features of the message (such as keywords, classification labels, etc.). The client behavior data can include the actual interaction behaviors of the client with these historical push messages, such as whether to open the message, reading duration, whether there is interaction (such as liking, commenting, forwarding, etc.). These behavior data reflect the user's preferences for different message types.

[0072] By combining the sample push data and the client behavior data, the message recognition model can comprehensively understand the interest patterns of the client in push messages from different service accounts.

[0073] In practical applications, when the target service account sends new push messages to the client, the instruction server can input these push messages into the already trained message recognition model. The message recognition model can calculate the preference matching degree of the client for the target service account based on the characteristics of the push messages and the historical behavior data of the user. The preference matching degree can be a quantitative index, for example, it can be expressed in the form of a percentage or a score, and is used to measure the degree of interest of the user in the service account or its push messages.

[0074] For example, if the message recognition model calculates that the preference matching degree of the user for the push messages of a certain clothing brand is only 10%, it indicates that the user is hardly interested in the push messages of this clothing brand and may need to be blocked. If the preference matching degree is as high as 80%, it indicates that the user is very interested in the push messages of this clothing brand and should be normally displayed.

[0075] 203: Generate a message filtering instruction containing the target identification information of the target service account when the preference matching degree meets the filtering condition.

[0076] In the embodiments of the present application, the instruction server can preset some filtering conditions for determining whether to block the push messages of the target service account. For example, the filtering conditions can include a preference matching degree threshold; when the preference matching degree is lower than the preference matching degree threshold, it can be determined that the preference matching degree meets the filtering condition, otherwise, it is determined that the preference matching degree does not meet the filtering condition. In addition, the filtering conditions can also include, for example, a message proportion limit. For example, when the proportion of the push messages of a certain service account in the total received push messages of the client is higher than a certain proportion, and the preference matching degree of the user for this service account is lower than the preference matching degree threshold, it is determined that the filtering condition is met, otherwise it is not met.

[0077] When the instruction server determines that the target service account meets the filtering condition according to the preference matching degree, a message filtering instruction containing the target identification information of the target service account can be generated.

[0078] 204: Send the message filtering instruction to the client in a silent push manner, so that when the target client receives at least one push message, it determines the service account corresponding to the at least one push message, and when the service account is the target service account, filters the at least one push message.

[0079] Among them, silent push is a special push method that does not display any notifications on the user's device but directly performs operations silently in the background. By using the silent push method to send the message filtering instruction to the client, the user will not perceive the existence of the message filtering instruction, avoiding unnecessary disturbances. Moreover, silent push can quickly transmit the message blocking instruction in the background without the user manually initiating a network request, greatly shortening the response time of the blocking operation.

[0080] After receiving the silent push, the client can first verify the integrity and legality of the message blocking instruction to ensure that the message blocking instruction has not been tampered with and complies with the compatibility standards of the current system.

[0081] After the client successfully verifies the message blocking instruction, it can parse the message blocking instruction to extract the target identification information of the target service account. The target identification information can be stored in the local database of the client for subsequent message comparison and filtering.

[0082] When the client receives a new push message, it can first check whether the service account corresponding to the message is the target service account. If the message source is the target service account, it will be directly intercepted and not shown to the user.

[0083] In the embodiment of the present application, by adopting the following steps: obtaining at least one push message sent by the target service account request to the client; using the message recognition model to determine the preference matching degree of the client for the target service account according to at least one push message, wherein the message recognition model is trained with sample push data and the sample behavior data of the client for the sample push data as input data, and the sample preference matching degree of the client for the sample service account corresponding to the sample push data as the training label; generating a message filtering instruction containing the target identification information of the target service account when the preference matching degree meets the filtering condition; sending the message filtering instruction to the client in the form of silent push, so that when the target client receives at least one push message, it can determine the service account corresponding to at least one push message, and when the service account is the target service account, perform filtering processing on at least one push message. The technical solution enables the instruction server to actively and accurately predict the client's interest level in the service account by using the message recognition model, thereby realizing the precise shielding of push messages, reducing the frequent interaction between the client and the server, and reducing network load and resource consumption. In addition, with the help of silent push technology, the blocking instruction can be sent to the client background without manual operation by the user, significantly improving the response speed.

[0084] In some embodiments, each of the at least one push messages has a corresponding message type.

[0085] In the embodiments of the present application, each push message may have a corresponding message type, which can be used to classify and identify push messages. For example, in an e-commerce application, the message types may include promotional advertisements, new product recommendations, system notifications, etc.

[0086] In some embodiments, using a message recognition model, determining the preference matching degree of the client for the target service account according to at least one push message can be specifically implemented as follows:

[0087] For any push message, determine the message preference matching degree of the client for the push message through the message recognition model.

[0088] For any push message, instruct the server to input it into the message recognition model. The message recognition model can calculate the message preference matching degree of the user for the push message according to the historical behavior data of the client and the characteristics of the message itself.

[0089] In some embodiments, generating a message filtering instruction including the target identification information of the target message source can be specifically implemented as follows:

[0090] When the push message meets the filtering condition, determine the message type corresponding to the push message;

[0091] Generate a message filtering instruction according to the message type and the target identification information, so that when the target client receives at least one push message from the target service account corresponding to the target identification information, filter the target push message belonging to the message type in the at least one push message.

[0092] When the server instructs to determine that a certain push message needs to be blocked, it can determine the message type of the push message. Then, the server can generate a message filtering instruction according to the message type and the target identification information of the target service account.

[0093] When the client receives a new push message, it can first check whether the service account corresponding to the push message is the target service account. If the message source is the target service account, the client can further compare whether the message type of the message is consistent with the type specified in the message filtering instruction. If this push message is both from the target service account and belongs to the specified message type, directly intercept the message and no longer display it to the user.

[0094] For example, in an e-commerce application, if a user often opens and reads the new product recommendation messages of a certain clothing brand for a long time, but hardly reads the promotional advertisements of this brand, the message recognition model can determine that the user has a low interest in the promotional advertisements of this brand and may need to block such messages. At this time, a message filtering instruction containing the message type of "promotional advertisement" can be generated and sent to the client through silent push. After receiving the message filtering instruction, if the client receives a push message of the promotional advertisement type again when the clothing brand sends it, the client can directly intercept the push message to avoid interfering with the user. In this way, personalized and more fine-grained filtering of push messages from service accounts can be achieved.

[0095] In some embodiments, the method may further include:

[0096] Obtaining the sample push messages received by the client within the first historical period, and the sample behavior data of the client for the sample push messages, where the sample behavior data includes one or more of the message opening frequency, message reading duration, interaction behavior, and message content;

[0097] Obtaining the sample preference matching degree of the client for the sample service account corresponding to the sample push message;

[0098] Training a message recognition model using the sample push messages, sample behavior data, and sample preference matching degree.

[0099] In order to train a message recognition model that can accurately judge the client's preference for messages, the instruction server needs to obtain sample push messages and sample behavior data. The sample push messages and sample behavior data can come from the messages received by the client within the first historical period (such as the past month) and their corresponding user behavior records.

[0100] Among them, the message opening frequency can refer to the number of times the client opens the push message or the proportion of the number of times the push message is opened to the total number of received times. For example, if a user often opens the promotional advertisements of a certain brand, it indicates an interest in such messages.

[0101] The message reading duration can refer to the length of time the user reads a certain push message through the client. For example, reading a message for a long time may indicate that the user is interested in the message content.

[0102] The interaction behavior can refer to whether the client has performed operations such as liking, commenting, and forwarding on the message. These behaviors are a direct manifestation of the user's interest.

[0103] After obtaining the sample push messages and sample behavior data, the instruction server will further obtain the sample preference matching degree of the user for the sample service accounts corresponding to these messages. The sample preference matching degree is a quantitative index used to measure the overall interest degree of the user in a certain service account or its push messages.

[0104] After obtaining the sample push messages, sample behavior data, and sample preference matching degree, the instruction server can use these data to train the message recognition model.

[0105] In some embodiments, training the message recognition model using the sample push messages, sample behavior data, and sample preference matching degree can be specifically implemented as follows:

[0106] Input the sample push messages and sample behavior data into the message recognition model to be trained, and the message recognition model to be trained outputs the predicted matching degree of the client for the sample service account;

[0107] According to the sample preference matching degree and the predicted matching degree, adjust the network parameters of the message recognition model to be trained to obtain the message recognition model.

[0108] After obtaining the sample push messages, sample behavior data, and sample preference matching degree, the instruction server can use these data to train the message recognition model with the goal of learning to predict the preference matching degree of the user for the service account according to the actual behavior data and message characteristics of the client.

[0109] In a possible implementation manner of the present application, the random forest classification algorithm can be used for model training.

[0110] In the initial stage, the historical data in the first historical period can be used for offline training to generate a preliminary model. After that, as the user continuously generates new message interaction behaviors during daily use, the instruction server can collect these new data in real time and perform incremental updates on the model at daily time intervals. For example, when the amount of data for a new day reaches 100, the instruction server will automatically trigger the model update process to ensure that the model can adapt to the dynamic changes of the user's interests in a timely manner.

[0111] In some embodiments, after sending the message filtering instruction to the client in the form of silent push, the method may further include:

[0112] Obtain the historical push messages received by the client in the second historical period and the historical behavior data of the historical push messages;

[0113] Construct incremental training data according to the historical push messages and historical behavior data;

[0114] Use the incremental training data to perform incremental training on the message preference model.

[0115] After obtaining the historical push messages and historical behavior data within the second historical period, the instruction server can integrate this data into incremental training data. Since the incremental training data is derived from the client's recent behavior, it can capture the dynamic changes in the client's interests. Therefore, the incremental training data can be used to supplement and update the training set of the message preference model, enabling it to reflect the user's latest interest changes.

[0116] After constructing the incremental training data, the instruction server can use the incremental training data to perform incremental training on the message preference model. Incremental training is an efficient way to update the model. It can quickly absorb the information in the new data while retaining the knowledge of the original model, thereby improving the prediction accuracy of the model.

[0117] During the process of performing incremental training on the model, the incremental training data can be input into the existing message preference model for fine-tuning and optimization of the model parameters. For example, the random forest algorithm can adapt to the new data by adding new decision trees or adjusting the weights of the existing decision trees. After the incremental training is completed, the updated model can be evaluated using the validation set and the test set to ensure that its performance meets the expected standards. For example, check whether the model can accurately predict the user's interest level in new messages.

[0118] In the embodiments of this application, the message preference model is trained and generated through machine learning techniques and is used to judge the user's interest level in different message types. To ensure that the client always uses the latest shielding rules, the system assigns a model version number to the message preference model. As the user behavior data accumulates continuously, the message preference model will be incrementally trained and optimized regularly, thereby generating new filtering rules.

[0119] In some embodiments, the method may further include:

[0120] Determine the current model version of the message preference model;

[0121] In some embodiments, specifically implementing the generation of a message filtering instruction including the target identification information of the target service account may be achieved as follows:

[0122] Determine the version number of the current version of the message recognition model;

[0123] Generate a message filtering instruction based on the version number and the target identification information, so that the client can update the filtering rules according to the version number. The filtering rules are used to record the identification information corresponding to the service accounts that need to be filtered.

[0124] To ensure that the client always uses the latest filtering rules, the instruction server can assign a model version number to the message preference model. The version number can be implemented as a three-digit code (such as 001, 002, etc.), and it will increment each time the model is updated. Through the version number, the client can quickly determine whether the received message filtering instruction is the latest version, thus avoiding mis-blocking or missed blocking caused by inconsistent rules. For example, if after an incremental training of the message preference model, it is found that the user's interest in a certain type of message has changed, a new version number will be generated (such as updated from 001 to 002). This new version number will be embedded in the subsequent generated message filtering instructions to ensure that the client can update the local blocking rules in a timely manner.

[0125] After receiving the message filtering instruction, the client can first check whether the version number in the message filtering instruction is higher than the version stored locally. Only when the verification passes and the version is the latest will subsequent processing be carried out. This mechanism ensures that the client always uses the latest filtering rules and avoids mis-blocking or missed blocking caused by old version rules.

[0126] After the client successfully verifies the message filtering instruction, it can store the target identification information and version number of the target service account in the local database. These information constitute the filtering rules of the client and are used to record the service accounts that need to be filtered.

[0127] Figure 3 The flowchart of a message push method provided by another embodiment of the present application is shown. This method can be applied to the client, such as Figure 3 As shown, the method may specifically include the following steps:

[0128] 301: Receive the first message filtering instruction sent by the server in a silent push manner. The first message filtering instruction contains the target identification information of the target service account. Among them, the preference matching degree of the client for the target service account is lower than the preset threshold. The preference matching degree is determined by the message recognition model. The message recognition model is trained with sample push data and the client's sample behavior data for the sample push data as input data, and the sample preference matching degree of the client for the sample service account corresponding to the sample push data as the training label.

[0129] 302: Receive the push message to be processed;

[0130] 303: Determine the first identification information of the service account corresponding to the push message to be processed;

[0131] 304: Determine whether the target identification information matches the first identification information;

[0132] 305: If the target identification information matches the first identification information, filter the push message to be processed.

[0133] In some embodiments, the method may further include:

[0134] Determine the first version number included in the first message filtering instruction, where the first version number is determined based on the current version of the message recognition model;

[0135] Determine the second version number included in the second message filtering instruction locally stored on the client;

[0136] In the case where the first version number is higher than the second version number, update the filtering rules stored locally using the target identification information included in the first message filtering instruction, where the filtering rules are used to record identification information.

[0137] Figure 3 The implementation process of the shown message push method can refer to Figure 2 the shown message push method, which will not be elaborated here.

[0138] Figure 4 The block diagram of a message push device provided by an embodiment of the present application is shown. The device can be applied to an instruction server, such as Figure 4 as shown, the device may specifically include:

[0139] A message acquisition module 401, configured to acquire at least one push message sent by a target service account to a client;

[0140] A preference determination module 402, configured to use a message recognition model to determine the preference matching degree of the client for the target service account according to at least one push message, where the message recognition model is trained with sample push data and sample behavior data of the client for the sample push data as input data, and the sample preference matching degree of the client for the sample service account corresponding to the sample push data as the training label;

[0141] An instruction generation module 403, configured to generate a message filtering instruction including the target identification information of the target service account when the preference matching degree meets the filtering condition;

[0142] An instruction push module 404, configured to send the message filtering instruction to the client in a silent push manner, so that when the target client receives at least one push message, determine the service account corresponding to at least one push message, and perform filtering processing on at least one push message when the service account is the target service account.

[0143] In some embodiments, at least one push message has a corresponding message type;

[0144] In some embodiments, the preference determination module 402 is specifically configured to:

[0145] For any push message, determine the message preference matching degree of the client for the push message through the message recognition model;

[0146] In some embodiments, the instruction generation module 403 is specifically configured to:

[0147] When the push message meets the filtering condition, determine the message type corresponding to the push message;

[0148] Generate a message filtering instruction according to the message type and the target identification information, so that when the target client receives at least one push message from the target service account corresponding to the target identification information, filter the target push messages belonging to the message type in the at least one push message.

[0149] In some embodiments, the device may further include:

[0150] A sample message acquisition module, configured to acquire the sample push messages received by the client within the first historical period, and the sample behavior data of the client for the sample push messages, where the sample behavior data includes one or more of the message opening frequency, message reading duration, interaction behavior, and message content;

[0151] A label acquisition module, configured to acquire the sample preference matching degree of the client for the sample service account corresponding to the sample push message;

[0152] A model training module, configured to train a message recognition model by using the sample push messages, the sample behavior data, and the sample preference matching degree.

[0153] In some embodiments, the model training module is specifically configured to:

[0154] Input the sample push messages and the sample behavior data into the message recognition model to be trained, and the message recognition model to be trained outputs the predicted matching degree of the client for the sample service account;

[0155] Adjust the network parameters of the message recognition model to be trained according to the sample preference matching degree and the predicted matching degree to obtain the message recognition model.

[0156] In some embodiments, the message push device further includes:

[0157] A historical message acquisition module, configured to acquire the historical push messages received by the client within the second historical period, and the historical behavior data for the historical push messages;

[0158] An incremental data generation module, configured to construct incremental training data according to the historical push messages and the historical behavior data;

[0159] An incremental training module, configured to perform incremental training on the message preference model by using the incremental training data.

[0160] In some embodiments, the message push device further includes:

[0161] A model version determination module, configured to determine the version number corresponding to the current model version of the message preference model;

[0162] In some embodiments, the instruction generation module 403 is specifically configured to:

[0163] Generate a message filtering instruction according to the version number and the target identification information, so that the client updates the filtering rule according to the version number, and the filtering rule is used to record the identification information corresponding to the service account that needs to be filtered.

[0164] In some embodiments, the message acquisition module 401 is specifically configured to:

[0165] Intercept at least one push message sent by the message server corresponding to the target service account to the client.

[0166] Figure 4 The described message push device can execute Figure 2 The message push method described in the illustrated embodiment, and its implementation principle and technical effects will not be elaborated. For the message push device in the above embodiments, the specific manners in which each module and unit perform operations have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0167] Figure 5 The block diagram of a message push device provided in another embodiment of the present application is shown. This device can be applied to a client, such as Figure 5 As shown, this device may specifically include:

[0168] An instruction receiving module 501, configured to receive a first message filtering instruction sent by the server in a silent push manner. The first message filtering instruction includes the target identification information of the target service account. Among them, the preference matching degree of the client for the target service account is lower than a preset threshold, and the preference matching degree is determined by a message recognition model. The message recognition model is trained with sample push data and the sample behavior data of the client for the sample push data as input data, and the sample preference matching degree of the client for the sample service account corresponding to the sample push data as the training label;

[0169] A message receiving module 502, configured to receive the push message to be processed;

[0170] An identification information determination module 503, configured to determine the first identification information of the service account corresponding to the push message to be processed;

[0171] A matching module 504, configured to determine whether the target identification information matches the first identification information;

[0172] A filtering module 505, which is used to filter the push message to be processed if so.

[0173] In some embodiments, the apparatus may further include:

[0174] A first version number determination module, which is used to determine a first version number included in the first message filtering instruction, and the first version number is determined based on the current version of the message recognition model;

[0175] A second version number determination module, which is used to determine a second version number included in the second message filtering instruction stored locally on the client;

[0176] A rule update module, which is used to update the filtering rules stored locally by using the target identification information included in the first message filtering instruction when the first version number is higher than the second version number, and the filtering rules are used to record identification information.

[0177] Figure 5 The described message push device may execute Figure 3 The message push method described in the illustrated embodiment, and its implementation principle and technical effects will not be elaborated. For the message push device in the above embodiment, the specific manners in which each module and unit perform operations have been described in detail in the embodiment related to the method, and will not be elaborated here.

[0178] An embodiment of the present application further provides a computing device, as Figure 6 shown, the device may include a storage component and a processing component;

[0179] The storage component stores one or more computer instructions, and among them, the one or more computer instructions are called by the processing component for execution to implement the message push method provided by the embodiment of the present application.

[0180] Of course, the computing device necessarily may further include other components, such as an input / output interface, a display component, a communication component, etc.

[0181] The input / output interface provides an interface between the processing component and the peripheral interface module, and the above peripheral interface module may be an output device, an input device, etc. The communication component is configured to facilitate communication between the computing device and other devices in a wired or wireless manner, etc.

[0182] Among them, the processing component may include one or more processors to execute computer instructions to complete all or part of the steps in the above methods. Of course, the processing component can also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components for executing the above methods.

[0183] The storage component is configured to store various types of data to support the operation of the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0184] The display component can be an electroluminescent (EL) element, a liquid crystal display or a microdisplay with a similar structure, or a retina-direct display or a similar laser scanning display.

[0185] It should be noted that the above computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. It can be implemented as a distributed cluster composed of multiple servers or terminal devices, or can be implemented as a single server or a single terminal device.

[0186] The embodiment of the present application also provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a computer, it can implement the message pushing method provided by the embodiment of the present application. The computer-readable medium can be included in the electronic device described in the above embodiment; or can exist alone without being assembled into the electronic device.

[0187] The embodiment of the present application also provides a computer program product, which includes a computer program carried on a computer-readable storage medium, and when the computer program is executed by a computer, it can implement the message pushing method provided by the embodiment of the present application. In such an embodiment, the computer program can be downloaded and installed from the network and / or installed from a removable medium. When the computer program is executed by a processor, it executes various functions defined in the system of the present application.

[0188] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described systems, devices and units can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0189] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative work.

[0190] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0191] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present application.

Claims

1. A message push method, characterized in that: Applied to the instruction server, the method includes: Obtain at least one push message that the target service account requests to send to the client; Determine the preference matching degree of the client to the target service account according to the at least one push message by using a message recognition model, wherein the message recognition model is generated by taking sample push data and sample behavior data of the client to the sample push data as input data, and taking the sample preference matching degree of the client to the sample service account corresponding to the sample push data as training labels; In the case where the preference matching degree satisfies the filtering condition, generating a message filtering instruction including the target identification information of the target service account; The message filtering instruction is sent to the client in a silent push manner, so that when the target client receives the at least one push message, it determines the service account corresponding to the at least one push message, and filters the at least one push message when the service account is the target service account.

2. The method according to claim 1, characterized in that The at least one push message has a corresponding message type respectively; The using a message identification model to determine the preference matching degree of the client to the target service account according to the at least one push message includes: For any push message, determining the message preference matching degree of the client to the push message through the message recognition model; The generating of the message filtering instruction including the target identification information of the target message source comprises: In the case where the push message meets the filtering condition, determining the message type corresponding to the push message; The message filtering instruction is generated according to the message type and the target identification information so that when the target client receives at least one push message from the target service account corresponding to the target identification information, the target push message belonging to the message type in the at least one push message is filtered and processed.

3. The method according to claim 2, characterized in that The method further comprises: Obtaining sample push messages received by the client in the first historical period, and sample behavior data of the client in response to the sample push messages, wherein the sample behavior data includes one or more of message opening frequency, message reading time, interactive behavior, and message content; Obtaining a sample preference matching degree of the client to the sample service account corresponding to the sample push message; The message recognition model is obtained by training using the sample push messages, sample behavior data and the sample preference matching degree.

4. The method according to claim 3, characterized in that The message recognition model is obtained by training the sample push message, the sample behavior data and the sample preference matching degree, including: Inputting the sample push message and sample behavior data into a message recognition model to be trained, and the message recognition model to be trained outputs a predicted matching degree of the client to the sample service account; According to the sample preference matching degree and the predicted matching degree, the network parameters of the message recognition model to be trained are adjusted to obtain the message recognition model.

5. The method according to claim 3, characterized in that: After sending the message filtering instruction to the client in a silent push manner, the method further includes: Acquire historical push messages received by the client in a second historical period, and historical behavior data for the historical push messages; Constructing incremental training data according to the historical push messages and the historical behavior data; The message preference model is incrementally trained using the incremental training data.

6. The method according to claim 5, characterized in that The method further comprises: Determining a version number corresponding to a current model version of the message preference model; The generating of the message filtering instruction including the target identification information of the target service account comprises: The message filtering instruction is generated according to the version number and the target identification information, so that the client updates the filtering rule according to the version number, and the filtering rule is used to record the identification information corresponding to the service account that needs to be filtered.

7. The method according to claim 1, characterized in that The at least one push message sent to the client by the target service account request includes: The at least one push message sent by the message server corresponding to the target service account to the client is intercepted.

8. A message push method, characterized in that: Applied to a client, the method comprises: Receive a first message filtering instruction sent by a server in a silent push mode, wherein the first message filtering instruction includes target identification information of a target service account, wherein a preference matching degree of the client to the target service account is lower than a preset threshold, and the preference matching degree is determined by a message recognition model, wherein the message recognition model is input with sample push data and sample behavior data of the client to the sample push data, and is generated by training with a sample preference matching degree of the client to a sample service account corresponding to the sample push data as a training label; Receive pending push messages; Determine first identification information of a service account corresponding to the push message to be processed; Determining whether the target identification information matches the first identification information; If so, the push message to be processed is filtered.

9. The method according to claim 8, characterized in that The method further comprises: Determine a first version number included in the first message filtering instruction, where the first version number is determined based on a current version of the message identification model; Determine a second version number included in a second message filtering instruction stored locally on the client; In the case where the first version number is higher than the second version number, the target identification information included in the first message filtering instruction is used to update the locally stored filtering rules, where the filtering rules are used to record identification information.

10. A computing device, characterized in that: including a processing component and a storage component; The storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the message pushing method as described in any one of claims 1 to 9.

11. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by the processing component, the message pushing method as described in any one of claims 1 to 9 is implemented.

12. A computer program product, characterized in that It includes a computer program / instruction, which, when executed by a processing component, implements the message push method as described in any one of claims 1 to 9.

Images