Message Push Method, Device, Computer Device and Storage Medium

Abstract

The present application relates to a message pushing method, apparatus, computer device and storage medium, and relates to the field of computer technology. The method includes: obtaining a set of target users; the set of target users includes a plurality of target users to whom messages are to be pushed; grouping the plurality of target users to obtain at least two user groups; one user group corresponds to one message pushing time period; and pushing messages to the target users in each user group in sequence according to the sequence of the at least two message pushing time periods. Through the above method, it is avoided that the same target user receives multiple message pushes in a short period of time, and the situation that the target user is frequently disturbed is avoided, thereby improving the user experience and the effect of message pushing.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of Internet technologies, and particularly to a message push method, apparatus, computer device, and storage medium. Background Art

[0002] The function of Push (message push) is to improve user activity and stickiness, increase user retention rate, and further improve product functions and user participation. Ways such as reminders of major domestic and international events, reminders of major events in the field, important message pushes of people or friends being followed, and message reminders of content of interest can guide users to open relevant application programs.

[0003] In the related art, for the multi-type content push of Push (message push), the common practice is that all services are sorted independently and pushed independently, and there is no association and control between services.

[0004] However, in the above message push method, there are push time conflicts between different types of content. After the types increase, the probability of push time conflicts increases significantly. The same user will receive multiple message pushes from the same application program in a short period of time, resulting in a poor message push effect and a decline in user experience. Summary of the Invention

[0005] The embodiments of the present application provide a message push method, apparatus, computer device, and storage medium, which can avoid the same target user receiving multiple message pushes in a short period of time, avoid the situation that the target user is frequently disturbed, thereby improving user experience and the message push effect. The technical solution is as follows:

[0006] On the one hand, a message push method is provided, and the method includes:

[0007] Obtain a set of target users; the set of target users includes multiple target users to be pushed with messages;

[0008] Group the multiple target users to obtain at least two user groups; one user group corresponds to one message push time period;

[0009] Push messages to the target users in each user group in sequence according to the sequence of the at least two message push time periods.

[0010] On the other hand, a message push apparatus is provided, and the apparatus includes:

[0011] A user acquisition module, configured to acquire a set of target users; the set of target users includes multiple target users to be pushed with messages;

[0012] A grouping module, configured to group multiple said target users to obtain at least two user groups; one said user group corresponds to one message push time period;

[0013] A message push module, configured to sequentially push messages to the target users in each of said user groups in the order of at least two said message push time periods.

[0014] In a possible implementation manner, said message push module includes:

[0015] A set acquisition sub-module, configured to acquire a set of messages to be pushed corresponding to each of at least two said user groups; the set of messages to be pushed contains the messages to be pushed corresponding to each of the target users in the user group;

[0016] A message determination sub-module, configured to determine, based on the sets of messages to be pushed corresponding to each of at least two said user groups, the target messages to be pushed corresponding to each of the target users in each of said user groups;

[0017] A message push sub-module, configured to sequentially push the corresponding target messages to be pushed to the target users in each of said user groups in the order of at least two said message push time periods.

[0018] In a possible implementation manner, said set acquisition sub-module is configured to,

[0019] Based on the message push time periods corresponding to each of said user groups, determine the message acquisition time periods corresponding to each of said user groups;

[0020] Determine, as the sets of messages to be pushed corresponding to each of said user groups, the sets composed of the messages to be pushed of each of the target users obtained by each of said user groups during the corresponding message acquisition time periods.

[0021] In a possible implementation manner, said message determination sub-module includes:

[0022] A click-through rate acquisition unit, configured to acquire the predicted true click-through rates of each of the first messages to be pushed included in the set of messages to be pushed; the first message to be pushed is the message to be pushed corresponding to a first target user in the set of messages to be pushed; the first target user is any one of the target users included in the user group;

[0023] A message determination unit, configured to determine the top p first messages to be pushed with the highest predicted true click-through rates as the target messages to be pushed corresponding to the first target user, where p is a positive integer.

[0024] In a possible implementation, the click-through rate acquisition unit is configured to

[0025] obtain the predicted click-through rate of the first message to be pushed through a predicted click-through rate model;

[0026] determine the predicted true click-through rate of the first message to be pushed based on the mapping relationship between the predicted click-through rate and the true click-through rate.

[0027] In a possible implementation, the click-through rate acquisition unit is configured to

[0028] perform at least two rounds of pre-pushing on the first message to be pushed to obtain the true click-through rates corresponding to each of the at least two rounds of pre-pushing;

[0029] determine the predicted true click-through rate of the first message to be pushed based on the true click-through rates corresponding to each of the at least two rounds of pre-pushing.

[0030] In a possible implementation, the click-through rate acquisition unit is configured to

[0031] obtain the historical true click-through rates of the push messages of the same type as the first message to be pushed;

[0032] determine the predicted true click-through rate of the first message to be pushed based on the historical true click-through rates of the push messages of the same type as the first message to be pushed.

[0033] On the other hand, a computer device is provided, which includes a processor and a memory. The memory stores at least one computer program, and the at least one computer program is loaded and executed by the processor to implement the above-mentioned message push method.

[0034] On the other hand, a computer-readable storage medium is provided, in which at least one computer program is stored, and the computer program is loaded and executed by a processor to implement the above-mentioned message push method.

[0035] On the other hand, a computer program product is provided, which includes at least one computer program, and the computer program is loaded and executed by a processor to implement the message push methods provided in the above various optional implementations.

[0036] The technical solution provided by this application may include the following beneficial effects:

[0037] The message pushing method provided by this application, during the process of message pushing, divides multiple target users to be pushed messages into at least two user groups according to the setting of the message pushing time period, and sequentially pushes messages to the target users in each user group according to the sequence of the message pushing time period, thereby avoiding the situation that the same target user receives multiple message pushes in a short period of time, avoiding the situation that the target user is frequently disturbed, thereby improving the user experience and the effect of message pushing.

[0038] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit this application. Brief Description of the Drawings

[0039] The accompanying drawings here are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with this application, and are used together with the specification to explain the principles of this application.

[0040] Figure 1 Shows a flowchart of the message pushing method provided by an exemplary embodiment of this application;

[0041] Figure 2 Shows a flowchart of the message pushing method provided by an exemplary embodiment of this application;

[0042] Figure 3 Shows a schematic diagram of the message pushing process provided by an exemplary embodiment of this application;

[0043] Figure 4 Shows a schematic diagram of the message pushing process shown by an exemplary embodiment of this application;

[0044] Figure 5 Shows a schematic diagram of the joint sorting shown by an exemplary embodiment of this application;

[0045] Figure 6 Shows a block diagram of the message pushing device shown by an exemplary embodiment of this application;

[0046] Figure 7 Is a structural block diagram of a computer device shown according to an exemplary embodiment. Detailed Description of the Specific Embodiment

[0047] Here, the exemplary embodiments will be described in detail, and the examples are shown in the accompanying drawings. When the following description relates to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with this application. On the contrary, they are only examples of devices and methods consistent with some aspects of this application as detailed in the appended claims.

[0048] Push (message push) refers to the active message push from the server corresponding to the application or third-party tool to the client. Users can see the Push message notifications through the notification bar. For example, when the mobile device's screen is locked, they can see the messages pushed by social applications; and they can click on the notification bar to launch the relevant application and display the page corresponding to the message notification within the application.

[0049] However, for social media, the types of messages that users can receive are diverse, and it is unacceptable to send all of them to users for the following reasons: First, the quantity is huge; second, the message quality is uneven; third, it increases the power consumption of the terminal, causes great interference to users, and conflicts with the power management mechanism of the terminal operating system. In related technologies, in the scenario of multi-type message push, in order to avoid the overall excessive quantity of push messages received by users and ensure the push scale of a single service, the following two message push methods are provided: 1. Each type of message is sorted and pushed independently, and there is no association or control between different types of services. However, in order to prevent a user from receiving too much content, a maximum push quota is set for a user, and this quota may be related to the application ecosystem and the restrictions of the terminal manufacturer. For example, terminal manufacturer A restricts that each person can receive at most N message pushes per day, so the application can push at most N messages to each user per day. However, this will cause the message types with more push messages in the morning to occupy more push quantities, and it cannot ensure that the messages of the message types that users are interested in are pushed. 2. On the basis of the above, a maximum push quota can be set for each message type to ensure that the push messages of each message type have the opportunity to be pushed. In this way, each type can push messages at the most appropriate time. For example, it is set that the maximum number of Push messages that a user can receive per day is N, including M updates from followed bloggers, I hot news, J social message reminders, etc., and each message type only uses its own quota; however, although the above scheme can ensure the push quantity of the push messages of each message type, there may be time conflicts between different message types, and in the scenario of a large number of message types, the probability of overlapping push times will also increase significantly, resulting in users receiving different types of message pushes from the same application within a short period of time, leading to a decline in the user experience.

[0050] Based on the technical problems existing in the above related technologies, the present application provides a message push method, which can avoid the possibility of users being pushed messages multiple times, randomly, and disorderly within a time window, avoid users being frequently disturbed, and thus improve the message push effect; Figure 1 The flowchart of the message push method provided by an exemplary embodiment of the present application is shown. This method can be executed by a computer device, and the computer device can be implemented as the server of an application program for message push, such asFigure 1 As shown, the message push method may include the following steps:

[0051] Step 110, obtaining a set of target users; the set of target users contains multiple target users to whom messages are to be pushed.

[0052] Step 120, grouping the multiple target users to obtain at least two user groups; one user group corresponds to one message push time period.

[0053] In an embodiment of the present application, a message push cycle may be set. In one message push cycle, multiple message push time periods may be divided. For example, one message push cycle is divided into m message push time periods, where m is a positive integer; based on the number of message push time periods, the multiple target users are grouped to obtain m user groups. Illustratively, if the message push cycle is 1 hour and each message push time period is divided every 10 minutes, then each hour may contain 6 message push time periods. At this time, the multiple target users may be divided into 6 user groups; or, the message push cycle may also be set to one day and each message push time period is divided every hour, then each day may contain 24 message push time periods. At this time, the multiple target users may be divided into 24 groups. Or, a do-not-disturb time period may be defined. For example, the time period between 23:00 and 7:00 is set as the do-not-disturb time period. Then, each day may contain 16 message push time periods. Therefore, the multiple target users are divided into 16 groups. It should be noted that the division of the message push cycle and the division of the message push time periods in the message push cycle may be set differently based on actual needs, and the present application does not limit this.

[0054] Optionally, when grouping the multiple target users, the multiple target users may be randomly divided into at least two user groups, and the number of target users in each user group may be the same or different.

[0055] Step 130, sequentially pushing messages to the target users in each user group according to the order of the at least two message push time periods.

[0056] In one message push time period, messages are pushed to the target users in the user group corresponding to this message push time period, so as to sequentially complete the message push to the target users in at least two user groups according to the order of the at least two message push time periods.

[0057] In an embodiment of the present application, since each user group corresponds to a different message push time period respectively, therefore, in one message push cycle, the target users in this user group will only be traversed once, that is, one message push is performed, so that while the target users can receive message pushes, the annoyance caused by frequent message pushes is avoided.

[0058] In summary, in the message push method provided by this application, during the message push process, multiple target users to be pushed are divided into at least two user groups according to the settings of the message push time period, and the target users in each user group are pushed messages in the order of the message push time period, thereby avoiding multiple message pushes received by the same target user in a short period of time, avoiding the situation that the target user is frequently disturbed, thereby improving the user experience and the message push effect.

[0059] In the scenario of multi-type message push, since different types correspond to different messages to be pushed, and the differences between different types of messages to be pushed are relatively large. For example, the messages to be pushed can include social messages, hot news, and interest messages, etc.; among them, social messages can include messages sent by followed users, establishing a temporary conversation, or messages sent by users with a friend relationship, etc.; hot news can include hot news in a vertical field, regional hot news, full-scale hot news, etc.; interest messages can include same-city messages, short video messages, etc.; when pushing messages to the same user, it is necessary to determine which type of message to be pushed is currently being pushed. Based on this, the embodiments of this application provide a message push method. Figure 2 The flowchart of the message push method provided by an exemplary embodiment of this application is shown. This method can be executed by a computer device, and the computer device can be implemented as a server of an application program for message push, such as Figure 2 As shown, the message push method can include the following steps:

[0060] Step 210, obtain a set of target users; the set of target users contains multiple target users to be pushed messages.

[0061] Step 220, group multiple target users to obtain at least two user groups; one user group corresponds to one message push time period.

[0062] Step 230, obtain a set of messages to be pushed corresponding to each of the at least two user groups; the set of messages to be pushed contains the messages to be pushed corresponding to each target user in the user group.

[0063] Among them, the set of messages to be pushed can include multiple messages to be pushed corresponding to each target user respectively. That is to say, the same target user can correspond to multiple messages to be pushed, and the number of messages to be pushed corresponding to different target users can be different. Schematically, a set of messages to be pushed can include a messages to be pushed. Among them, target user 1 corresponds to b messages to be pushed, target user 2 corresponds to c messages to be pushed, target user 3 corresponds to d messages to be pushed, and so on. At this time, a = b + c + d +...; a, b, c, and d are all positive integers.

[0064] Since the messages to be pushed usually have timeliness, in order to ensure the timeliness of the messages to be pushed and avoid pushing messages with a long time interval to users, which may affect the message pushing effect, in a possible implementation, the set of messages to be pushed is the messages to be pushed of each target user collected during the message acquisition period corresponding to the user group. Based on this, the process of obtaining the sets of messages to be pushed corresponding to at least two user groups can be implemented as follows:

[0065] Based on the message pushing periods corresponding to each user group, determine the message acquisition periods corresponding to each user group respectively;

[0066] Determine the set composed of the messages to be pushed of each target user obtained by each user group during the corresponding message acquisition period as the set of messages to be pushed corresponding to each user group respectively.

[0067] Optionally, after obtaining the message pushing periods corresponding to each user group, the message acquisition periods corresponding to each user group can be determined based on the timeliness duration of the messages to be pushed; among them, the timeliness duration of the messages to be pushed can be set by relevant personnel based on actual needs. For example, the timeliness duration can be set to 24 hours, 12 hours, 6 hours, etc.

[0068] Optionally, the message pushing time can be included in the message pushing period, and the message pushing time can be any specified time in the message pushing period; when determining the message acquisition period, the message acquisition period can be determined based on the message pushing time in the message pushing period and in combination with the timeliness duration of the messages to be pushed.

[0069] Schematically, taking the message push cycle as 1 hour, dividing the message push period into units of 10 minutes, and the message aging period as 24 hours as an example. If the first message push period is from 7:00 to 7:10, and if the message is calculated and pushed at 7:00 (message push time), the message acquisition period corresponding to the first message push period is from 7:00 yesterday to 7:00 today; if the message is calculated and pushed at 7:10 (message push time), the message acquisition period corresponding to the first message push period is from 7:10 yesterday to 7:10 today; the set composed of the messages to be pushed corresponding to each target user in the user group collected during the first message push period is determined as the set of messages to be pushed corresponding to the user group.

[0070] Step 240, based on the sets of messages to be pushed corresponding to at least two user groups respectively, determine the target messages to be pushed corresponding to each target user in each user group.

[0071] Since the messages to be pushed of each type have different message contents, and only one message push is performed for the same user during the same message push period, it is necessary to determine the target messages to be pushed to the target user from the messages to be pushed of each type included in the current set of messages to be pushed. In order to make the messages to be pushed of multiple types comparable in the same spatial dimension, a unified index between the messages to be pushed of different types can be established, and the comparability between the messages to be pushed of multiple types can be realized by using this unified index, and this unified index is used to sort the messages to be pushed of multiple types to determine the target messages to be pushed.

[0072] Optionally, in the embodiment of the present application, the unified index is the predicted true click-through rate of each message to be pushed; based on this, taking any target user in any user group as an example, the process of determining the target messages to be pushed corresponding to each target user can be realized as follows:

[0073] Obtain the predicted true click-through rate of each first message to be pushed included in the set of messages to be pushed; the first message to be pushed is the message to be pushed corresponding to the first target user in the set of messages to be pushed; the first target user is any one of the target users included in the user group;

[0074] Determine the top p first messages to be pushed with the highest predicted true click-through rate as the target messages to be pushed for the first target user.

[0075] In the embodiments of the present application, since each target user corresponds to different messages to be pushed, and the push results of the same message to be pushed may also be different for different target users. Therefore, when predicting the true click-through rate of each message to be pushed, the calculation of predicting the true click-through rate can be performed separately for each target user, and the target messages to be recommended for each target user can be determined separately.

[0076] The target types of the first messages to be pushed corresponding to the first target user may be different.

[0077] In the embodiments of the present application, after obtaining the predicted true click-through rates of each of the first messages to be pushed, the first messages to be pushed can be sorted based on the predicted true click-through rates of each of the first messages to be pushed. Based on the sorting result, the top p first messages to be pushed with the highest predicted true click-through rates are determined as the target messages to be pushed for the first target user; if sorted in descending order of the predicted true click-through rate, the first p messages to be pushed in the sorting are determined as the target messages to be pushed for the first target user; if sorted in ascending order of the predicted true click-through rate, the last p messages to be pushed in the sorting are determined as the target messages to be pushed for the first target user.

[0078] Among them, the value of p can be set by relevant personnel based on actual needs, and the present application does not limit this. By way of illustration, when the value of p is 1, it means that within a message push cycle, 1 target message to be pushed is pushed to the first target user. Taking one target user as an example, Figure 3 shows a schematic diagram of the message push process provided by an exemplary embodiment of the present application, as Figure 3 shown, different message senders 310 provide messages to be pushed to the same target user. Different message senders may correspond to different services and send different types of messages to be pushed; the computer device 320 collects various types of messages to be pushed sent by each message sender to the same target user, and after jointly sorting the various types of messages to be pushed during the message push period corresponding to the user group where the target user is located, the one with the highest predicted true click-through rate in the joint sorting is determined as the target message to be pushed for the target user and pushed to the target user.

[0079] In the embodiments of the present application, there are multiple ways to obtain the predicted true click-through rate. Taking the first message to be pushed as an example, the ways to obtain the predicted true click-through rate include:

[0080] Method 1: Obtain the estimated click-through rate of the first message to be pushed through an estimated click-through rate model;

[0081] Based on the mapping relationship between the estimated click-through rate and the true click-through rate, determine the predicted true click-through rate of the first message to be pushed.

[0082] Among them, the predicted click-through rate model can be a machine learning model trained based on user characteristics of sample users, message characteristics of sample messages, sender characteristics of message senders of sample messages, and true click-through rates of sample messages; therefore, when obtaining the predicted click-through rate of the first message to be pushed through the predicted click-through rate model, it includes:

[0083] Obtain the message characteristics of the first message to be pushed, the sender characteristics of the message sender of the first message to be pushed, and the user characteristics of the first target user;

[0084] Input the message characteristics of the first message to be pushed, the sender characteristics of the message sender of the first message to be pushed, and the user characteristics of the first target user into the predicted click-through rate model to obtain the predicted click-through rate of the first message to be pushed output by the predicted click-through rate model.

[0085] In the embodiment of the present application, a mapping relationship between the predicted click-through rate and the true click-through rate is established; among them, the mapping relationship can be established based on the corresponding relationship between the predicted click-through rate and the true click-through rate of the pushed messages.

[0086] Method 2: Perform at least two rounds of pre-pushing on the first message to be pushed to obtain the true click-through rates corresponding to each of the at least two rounds of pre-pushing;

[0087] Based on the true click-through rates corresponding to each of the at least two rounds of pre-pushing, determine the predicted true click-through rate of the first message to be pushed.

[0088] Among them, pre-pushing means that before a large-scale push of the first message to be pushed, a small-scale random push of the first message to be pushed is first performed; that is, the number of pre-pushed users corresponding to the first message to be pushed is much smaller than the number of target users corresponding to the first message to be pushed; for example, if the number of pushed users of the first message to be pushed is 20 million, before a large-scale push, the first message to be pushed is randomly pushed to 200,000 pushed users (i.e., pre-pushed users), and then the first message to be pushed is randomly pushed to 2 million pushed users. After that, based on the true click-through rates corresponding to each of the two rounds of pre-pushing, the predicted true click-through rate of the first message to be pushed is determined, and then based on this predicted true click-through rate, it is determined whether to push the first message to be pushed to the remaining 17.8 million pushed users (target users); that is, during pre-pushing, the first message to be pushed is directly pushed. After the number of pre-pushing times reaches the set number of pushing times, based on the true click-through rates obtained from pre-pushing, the subsequent push of the first message to be pushed is guided. It should be noted that the proportion of the number of pre-pushed users in the total number of pushed users can be set by relevant personnel based on actual needs, and the proportions corresponding to different rounds of pre-pushing can be different.

[0089] Since both the pre-push and large-scale push of the first message to be pushed (i.e., sorting based on the predicted true click-through rate and determining whether to push) are instant pushes of the first message to be pushed, the pre-push process usually occurs within a short period before the large-scale push process. Schematically, at least two rounds of pre-push processes can be completed within 10 to 20 minutes before the large-scale push process, and the true click-through rates corresponding to each of the at least two rounds of pre-pushes are collected.

[0090] In the embodiments of the present application, the predicted true click-through rate of the first message to be pushed can be the average value of the true click-through rates corresponding to each of the at least two rounds of pre-pushes, or it can also be the maximum value. The present application does not limit this.

[0091] Method 3: Obtain the historical true click-through rate of push messages of the same type as the first message to be pushed;

[0092] Based on the historical true click-through rate of the push messages of the same type as the first message to be pushed, determine the predicted true click-through rate of the first message to be pushed.

[0093] Optionally, the computer device can determine the predicted true click-through rate of the first message to be pushed based on the historical true click-through rate of push messages of the same type.

[0094] In the embodiments of the present application, the message types of the messages to be pushed can include one of the first type, the second type, and the third type; among them, the push messages of the first type can be messages with weak timeliness, such as social messages, interest messages, etc.; the push messages of the second type can be messages with strong timeliness but whose push quality cannot be determined, such as hot messages in the entertainment field; the push messages of the third type can be messages with strong timeliness and guaranteed push quality, such as official regional hot messages, or messages sent to specific types of users, etc.; the push messages of the third type can correspond to a relatively small set of push users, that is, a segmented population; the segmented population refers to the population divided based on the login frequency. Users with similar login frequencies will have relatively consistent operations when facing different messages. It should be noted that the above examples of the types of social messages, interest messages, and hot messages are only schematic, and the present application does not limit the message content included in each message type.

[0095] In the embodiments of the present application, different message types correspond to different methods for determining the predicted true click-through rate.

[0096] Optionally, when the message type of the first message to be pushed is the first type, Method 1 is used to obtain the predicted true click-through rate of the first message to be pushed.

[0097] When the message type of the first message to be pushed is the second type, method two is used to obtain the predicted true click-through rate of the first message to be pushed.

[0098] When the message type of the first message to be pushed is the third type, method three is used to obtain the predicted true click-through rate of the first message to be pushed.

[0099] In the embodiments of the present application, by setting a unified index, the push messages of different types are comparable, so that in the same message push period, when there are multiple types of push messages for the same target user, the target message to be pushed to the target user this time can be determined.

[0100] Step 250: Push the corresponding target message to be pushed to the target users in each user group in sequence according to the order of at least two message push periods.

[0101] Taking the message push cycle as 1 hour, dividing each 10 minutes into a message push period, and the number of target users in each user group being the same as an example, Figure 4 shows a schematic diagram of the message push process shown in an exemplary embodiment of the present application, as Figure 4 shown, each 10 minutes corresponds to a message push period, and each message push period corresponds to a user group. In message push period 1 (for example, from 8:00 to 8:10), the target messages to be pushed for 1 / 6 of the target users included in user group 1 are determined by the joint sorting of multiple types of messages, and the corresponding target messages to be pushed are pushed to these 1 / 6 of the target users; in message push period 2 (for example, from 8:10 to 8:20), the target messages to be pushed for 1 / 6 of the target users included in user group 2 are determined by the joint sorting of multiple types of messages, and the corresponding target messages to be pushed are pushed to these 1 / 6 of the target users; and so on, until the message push for 1 / 6 of the target users included in user group 6 is completed in message push period 6; then, the next message push cycle is entered.

[0102] In the embodiments of the present application, the maximum push quota of users is shared among different types of messages. Figure 5 shows a schematic diagram of the joint sorting shown in an exemplary embodiment of the present application, as Figure 5 shown, taking the same target user as an example, different services (message senders) send their respective messages to be pushed to the target user. Among them, the message types sent by different services are different, as Figure 5As shown, it may include message-based services, hotspot-based services, social-based services, etc. After the computer device obtains the messages to be pushed separately sent by each service, it determines the predicted true click-through rates of the messages to be pushed of each message type through the determination method of the predicted true click-through rate corresponding to each message type. Through a multi-type mixed sorting model, it mixes and sorts the multi-type messages to be pushed based on the predicted true click-through rates of each message to be pushed, determines the q messages to be pushed with the highest predicted true click-through rates among them as the target messages to be pushed, and pushes them to the corresponding target users. At this time, multiple types of messages share the maximum push quota of the user, and a message push is performed on the target user once per message push cycle.

[0103] In summary, in the message push method provided by this application, during the message push process, by dividing multiple target users to be pushed messages into at least two user groups according to the settings of the message push time periods, and sequentially pushing messages to the target users in each user group according to the sequence of the message push time periods, it avoids the situation that the same target user receives multiple message pushes in a short period of time, avoids the situation that the target user is frequently disturbed, thereby improving the user experience and the message push effect.

[0104] At the same time, in the multi-type message push scenario, the embodiment of this application realizes the joint sorting of multi-type messages to be pushed in the same dimension by setting a unified index (i.e., the predicted true click-through rate), thereby effectively aggregating the messages to be pushed sent to the user, and selecting several messages to be pushed with the highest predicted true click-through rates from them for pushing, so that the messages pushed to the user are more in line with the user's needs, and improves the push quality of the messages, thereby improving the message push effect.

[0105] Figure 6 shows a block diagram of a message push device shown in an exemplary embodiment of this application. This message push device is used to implement all or part of the steps of the message push method shown in Figure 1 or Figure 2 any of the embodiments. As shown in Figure 6 this message push device may include:

[0106] A user acquisition module 610, configured to acquire a set of target users; the set of target users contains multiple target users to be pushed messages;

[0107] A grouping module 620, configured to group the multiple target users to obtain at least two user groups; one user group corresponds to one message push time period;

[0108] The message push module 630 is used to sequentially push messages to the target users in each of the user groups in the order of the at least two message push time periods.

[0109] In a possible implementation, the message push module 630 includes:

[0110] A set acquisition sub-module, configured to acquire a set of messages to be pushed corresponding to each of the at least two user groups; the set of messages to be pushed includes the messages to be pushed corresponding to each of the target users in the user group.

[0111] A message determination sub-module, configured to determine, based on the sets of messages to be pushed corresponding to each of the at least two user groups, the target messages to be pushed corresponding to each of the target users in each of the user groups.

[0112] A message push sub-module, configured to sequentially push the corresponding target messages to be pushed to the target users in each of the user groups in the order of the at least two message push time periods.

[0113] In a possible implementation, the set acquisition sub-module is configured to,

[0114] Based on the message push time periods corresponding to each of the user groups, determine the message acquisition time periods corresponding to each of the user groups;

[0115] Determine, as the set of messages to be pushed corresponding to each of the user groups, the set composed of the messages to be pushed of each of the target users acquired by each of the user groups during the corresponding message acquisition time period.

[0116] In a possible implementation, the message determination sub-module includes:

[0117] A click-through rate acquisition unit, configured to acquire the predicted true click-through rate of each first message to be pushed included in the set of messages to be pushed; the first message to be pushed is the message to be pushed corresponding to a first target user in the set of messages to be pushed; the first target user is any one of the target users included in the user group.

[0118] A message determination unit, configured to determine the top p first messages to be pushed with the highest predicted true click-through rate as the target messages to be pushed corresponding to the first target user, where p is a positive integer.

[0119] In a possible implementation, the click-through rate acquisition unit is configured to,

[0120] Obtain the estimated click-through rate of the first message to be pushed through an estimated click-through rate model;

[0121] Based on the mapping relationship between the estimated click-through rate and the true click-through rate, determine the predicted true click-through rate of the first message to be pushed.

[0122] In a possible implementation, the click-through rate acquisition unit is used to

[0123] Perform at least two rounds of pre-pushing on the first message to be pushed, and obtain the true click-through rate corresponding to each of the at least two rounds of pre-pushing;

[0124] Based on the true click-through rate corresponding to each of the at least two rounds of pre-pushing, determine the predicted true click-through rate of the first message to be pushed.

[0125] In a possible implementation, the click-through rate acquisition unit is used to

[0126] Obtain the historical true click-through rate of the push messages of the same type as the first message to be pushed;

[0127] Based on the historical true click-through rate of the push messages of the same type as the first message to be pushed, determine the predicted true click-through rate of the first message to be pushed.

[0128] In summary, the message push device provided by this application, during the message push process, divides multiple target users to be pushed into at least two user groups according to the settings of the message push time period, and sequentially pushes messages to the target users in each user group according to the order of the message push time period, thereby avoiding the situation that the same target user receives multiple message pushes within a short period of time, avoiding the situation that the target user is frequently disturbed, thereby improving the user experience and the effect of message push.

[0129] At the same time, in the multi-type message push scenario, the embodiments of this application realize the joint sorting of multi-type messages to be pushed in the same dimension by setting a unified index (i.e., the predicted true click-through rate), thereby effectively aggregating the messages to be pushed to the user, and selecting several messages to be pushed with the highest predicted true click-through rate for pushing, so that the messages pushed to the user more meet the user's needs, and improve the push quality of the messages, thereby improving the effect of message push.

[0130] Figure 7The structure block diagram of a computer device 700 shown in an exemplary embodiment of the present application is shown. The computer device may be implemented as the cloud terminal management platform in the above solution of the present application or a cloud server with cloud terminals built thereon. The computer device 700 includes a central processing unit (CPU) 701, a system memory 704 including a random access memory (RAM) 702 and a read-only memory (ROM) 703, and a system bus 705 connecting the system memory 704 and the central processing unit 701. The computer device 700 further includes a mass storage device 706 for storing an operating system 709, application programs 710 and other program modules 711.

[0131] Without loss of generality, the computer-readable medium may include a computer storage medium and a communication medium. The computer storage medium includes volatile and non-volatile, removable and non-removable media implemented by any method or technology for storing information such as computer-readable instructions, data structures, program modules or other data. The computer storage medium includes RAM, ROM, erasable programmable read-only registers (EPROM), electrically-erasable programmable read-only memories (EEPROM), flash memory or other solid-state storage technologies, CD-ROMs, digital versatile discs (DVDs) or other optical storage, magnetic tape cartridges, tapes, magnetic disk storage or other magnetic storage devices. Of course, those skilled in the art will understand that the computer storage medium is not limited to the above several types. The above system memory 704 and mass storage device 706 may be collectively referred to as the memory.

[0132] According to various embodiments of the present application, the computer device 700 may also run on a remote computer on the network through a network such as the Internet. That is, the computer device 700 may be connected to a network 708 through a network interface unit 707 connected to the system bus 705, or in other words, the network interface unit 707 may also be used to connect to other types of networks or remote computer systems (not shown).

[0133] The memory further includes at least one instruction, at least one program, a code set or an instruction set, which are stored in the memory. The central processing unit 701 implements all or part of the steps in the message pushing method shown in the above various embodiments by executing the at least one instruction, at least one program, the code set or the instruction set.

[0134] In an exemplary embodiment, a computer-readable storage medium is further provided. At least one computer program is stored in the computer-readable storage medium, and the computer program is loaded and executed by a processor to implement all or part of the steps in the above message pushing method. For example, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, a floppy disk, and an optical data storage device, etc.

[0135] In an exemplary embodiment, a computer program product is further provided. The computer program product includes at least one computer program, and the computer program is loaded and executed by a processor to implement Figure 1 or Figure 2 all or part of the steps in the message pushing method shown in any of the above embodiments.

[0136] Those skilled in the art will readily conceive of other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any variations, uses, or adaptations of the present application, which follow the general principles of the present application and include common general knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and embodiments are only to be considered as exemplary, and the true scope and spirit of the present application are pointed out by the following claims.

[0137] It should be understood that the present application is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present application is only limited by the appended claims.

Claims

1. A message push method, characterized in that, The method includes: Obtaining a set of target users; the set of target users includes multiple target users to whom message push is to be performed; Grouping the multiple target users to obtain at least two user groups; one user group corresponds to one message push time period; Sequentially pushing messages to the target users in each of the user groups according to the sequence of the at least two message push time periods; Among them, the sequentially pushing messages to the target users in each of the user groups according to the sequence of the at least two message push time periods includes: Obtaining a set of messages to be pushed corresponding to each of the at least two user groups; the set of messages to be pushed includes the messages to be pushed corresponding to each of the target users in the user group; Based on the sets of messages to be pushed corresponding to each of the at least two user groups, determining, for each of the target users in each of the user groups, the target message to be pushed corresponding to each of them; Sequentially pushing the corresponding target messages to be pushed to the target users in each of the user groups according to the sequence of the at least two message push time periods.

2. The method according to claim 1, wherein The obtaining the sets of messages to be pushed corresponding to each of the at least two user groups includes: Based on the message push time periods corresponding to each of the user groups, determining the message acquisition time periods corresponding to each of the user groups; Determining, as the set of messages to be pushed corresponding to each of the user groups, the set composed of the messages to be pushed of each of the target users obtained by each of the user groups during the corresponding message acquisition time period.

3. The method according to claim 1, wherein The determining, for each of the target users in each of the user groups, the target message to be pushed corresponding to each of them based on the sets of messages to be pushed corresponding to each of the at least two user groups includes: Obtaining the predicted true click-through rate of each first message to be pushed included in the set of messages to be pushed; the first message to be pushed is the message to be pushed corresponding to a first target user in the set of messages to be pushed; the first target user is any one of the target users included in the user group; Determining the top p first messages to be pushed with the highest predicted true click-through rate as the target messages to be pushed corresponding to the first target user, where p is a positive integer.

4. The method according to claim 3, characterized in that, The obtaining the predicted true click-through rate of each first message to be pushed included in the set of messages to be pushed includes: Obtaining the estimated click-through rate of the first message to be pushed through an estimated click-through rate model; Based on the mapping relationship between the estimated click-through rate and the true click-through rate, determining the predicted true click-through rate of the first message to be pushed.

5. The method according to claim 3, characterized in that, The obtaining the predicted true click-through rate of each first message to be pushed included in the set of messages to be pushed includes: Performing at least two rounds of pre-pushing on the first message to be pushed to obtain the true click-through rate corresponding to each of the at least two rounds of pre-pushing; Based on the true click-through rates corresponding to each of the at least two rounds of pre-pushing, determining the predicted true click-through rate of the first message to be pushed.

6. The method according to claim 3, wherein Obtaining the predicted true click-through rate of each first message to be pushed included in the set of messages to be pushed, includes: Obtaining the historical true click-through rate of push messages of the same type as the first message to be pushed; Based on the historical true click-through rate of the push messages of the same type as the first message to be pushed, determining the predicted true click-through rate of the first message to be pushed.

7. A message pushing device, characterized in that, The device includes: A user acquisition module, configured to acquire a set of target users; the set of target users includes multiple target users to whom messages are to be pushed; A grouping module, configured to group the multiple target users to obtain at least two user groups; one user group corresponds to one message push time period; A message push module, configured to sequentially push messages to the target users in each of the user groups in the order of the at least two message push time periods; Wherein, the message push module includes: A set acquisition sub-module, configured to acquire the set of messages to be pushed corresponding to each of the at least two user groups; the set of messages to be pushed includes the messages to be pushed corresponding to each of the target users in the user group; A message determination sub-module, configured to determine, based on the sets of messages to be pushed corresponding to each of the at least two user groups, the target messages to be pushed corresponding to each of the target users in each of the user groups; A message push sub-module, configured to sequentially push the corresponding target messages to be pushed to the target users in each of the user groups in the order of the at least two message push time periods.

8. A computer device, characterized in that, The computer device includes a processor and a memory, and the memory stores at least one computer program, and the at least one computer program is loaded and executed by the processor to implement the message push method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one computer program, and the computer program is loaded and executed by a processor to implement the message push method according to any one of claims 1 to 6.

Images