Intelligent message pushing method and device, equipment and storage medium

Abstract

The invention discloses an intelligent message pushing method, apparatus and device, and a storage medium. The intelligent message pushing method comprises the steps of obtaining a resource occupation value of a pushing server at a current moment; comparing the resource occupation value with a preset first threshold value and a preset second threshold value; if the resource occupation value is smaller than the first threshold value, pushing a message to a user based on a first rule; if the resource occupancy value is greater than the first threshold value and less than or equal to the second threshold value, pushing a message to a user group based on a second rule; and if the resource occupation value is greater than the second threshold value, pushing a message to the user by adopting a default channel. Different pushing strategies are adopted according to the occupation condition of the server resources, the technical problems that a large number of server resources are occupied and the message pushing success rate is low in the existing message pushing process are solved, and the optimal matching between the server resources and the pushing effect is achieved.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to an intelligent message push method, apparatus, device, and storage medium. Background Technology

[0002] Push notifications are the most common form of interaction with users, with common message carriers including SMS, app messages, email, WeChat, and 5G messaging. Once a message transmission channel is established, operations personnel can transmit message content from the server to the user's terminal.

[0003] In existing technologies, the following two methods are commonly used for message push: The first type is traditional message delivery technology, such as Figure 1 As shown, this method relies on message queue technology components. After a batch message sending task is initiated, each message queues up to be sent. Logically, the packaged message tasks are arranged linearly. After sending is completed, a notification is sent to further complete the sending of the next queued message. For high load and high concurrency situations, an asynchronous sending mode can also be adopted, eliminating the need to wait for a successful sending notification and starting the sending of the next message according to a pre-defined strategy. While this message delivery technology can complete batch message sending tasks, considering user receiving channels, sending failures may occur, such as the user's number not existing, the app being deleted, or the email not being opened for a long time. In such cases, the message will not reach the user.

[0004] The second type is tiered push technology, such as... Figure 2 As shown, this method extracts each message from a batch push task and sends it through various channels such as APP messages, email, WeChat, and SMS according to the configured order until successful delivery. While this message delivery technology can guarantee message reach, it consumes huge system resources and cannot clearly define the preferred message sending method, resulting in significant resource waste. For example, one batch of users might have the highest reach rate via email, while the next batch might have the highest reach rate via APP messages. Determining the first message channel also creates operational challenges. In other words, tiered push has the dual drawbacks of high resource consumption and poor marketing effectiveness. On the one hand, tiered push requires additional processes to constantly monitor the failure rate of existing message channels, resulting in huge resource consumption. When a large number of messages are waiting to be sent, it affects the normal operation of other server services. On the other hand, considering that users have different levels of attention to different channels, successfully pushed messages may not be seen by users, which is equivalent to not sending them at all. Summary of the Invention

[0005] The present application aims to at least solve one of the technical problems in the related art. To this end, one object of the present application is to provide an intelligent message pushing method, device, equipment and storage medium, which can solve the technical problems of large server resource occupation and low message pushing success rate in the prior art, and realize different pushing strategies according to the server resource occupation, thereby more reasonably utilizing the server resources and improving the message pushing success rate.

[0006] The technical solution adopted by the present application is: In a first aspect, the present application provides an intelligent message pushing method, comprising: obtaining a resource occupation value of a pushing server at a current time; comparing the resource occupation value with a preset first threshold value and a preset second threshold value; if the resource occupation value is less than the first threshold value, pushing a message to a user based on a first rule; if the resource occupation value is greater than the first threshold value and less than or equal to the second threshold value, pushing a message to a user group based on a second rule; and if the resource occupation value is greater than the second threshold value, pushing a message to a user using a default channel.

[0007] The message pushing to the user based on the first rule comprises: calculating the adaptation degree of each channel c for the current user u using a first adaptation degree calculation formula; sorting the adaptation degrees of all channels according to the size, and taking the channel with the largest adaptation degree value as the pushing channel of the current user u; wherein the first adaptation degree calculation formula is: In the formula, is the adaptation degree of the current user u to the channel c, is the historical reach rate of the current user u to the channel c, is the activity of the reach target using the channel c, is the real-time occupation rate of the channel c, , , is the weight, , , The value range of is 0~1, wherein the types of the channel include: SMS, or APP, or WeChat, or email.

[0008] The message pushing to the user group based on the second rule comprises: calculating the adaptation degree of each channel c for the current user group using a second adaptation degree calculation formula, and the current user group is the user group with the same label as the current user u; sorting the adaptation degrees of all channels according to the size, and taking the channel with the largest adaptation degree value as the pushing channel of the current user group; wherein the second adaptation degree calculation formula is: In the formula, ​​​For the current user group The adaptation degree to each channel c, The number of current user groups, For the current user group The weight of the historical reach rate using channel c, The historical reach rate of channel C, For the current user group The historical adaptation degree of channel C, The sensitivity coefficient, For the current user group The channel adaptation degree threshold.

[0009] Wherein, the resource occupation value of the push server at the current time is obtained, including: obtaining the occupation rate of CPU, memory, and bandwidth of the push server; calculating the resource occupation value of the push server according to the resource occupation calculation formula; wherein, the resource occupation calculation formula is: In the formula, R is the resource occupation value of the push server, The occupation rate of CPU of the push server, The occupation rate of memory of the push server, The occupation rate of bandwidth of the push server, W i The weight corresponding to each occupation rate.

[0010] In a second aspect, the present application provides an intelligent message pushing device, comprising: a server resource occupation obtaining module for obtaining the resource occupation value of the push server at the current time; a resource occupation value comparison module for comparing the resource occupation value with a preset first threshold and a preset second threshold; a first message pushing module for pushing messages to users based on a first rule if the resource occupation value is less than the first threshold; a second message pushing module for pushing messages to a user group based on a second rule if the resource occupation value is greater than the first threshold and less than or equal to the second threshold; and a third message pushing module for pushing messages to users using a default channel if the resource occupation value is greater than the second threshold.

[0011] Wherein, the first message pushing module comprises: a first adaptation degree calculation unit for calculating the adaptation degree of each channel c for the current user u using a first adaptation degree calculation formula; and a first channel determination unit for sorting the adaptation degrees of all channels according to size and taking the channel with the largest adaptation degree value as the push channel of the current user u; wherein, the first adaptation degree calculation formula is: In the formula, The adaptation degree of the current user u using channel c, The historical reach rate of the current user u using channel c, The activity of the reach target using channel c, This represents the real-time occupancy rate of channel c. , , As weight, , , The value range is 0 to 1.

[0012] The second message push module includes a second adaptation calculation unit, used to calculate the current user group using a second adaptation calculation formula. The adaptation degree of each channel c; the second channel determination unit, used to sort all channels by their adaptation degree and select the channel with the highest adaptation degree as the current user group. The push channel; wherein, the formula for calculating the second adaptability is: In this formula, For the current user group The fit of each channel c, The number of users in the current user group. For the current user group The weighting of historical reach rates using channel c This represents the historical reach rate of channel C. For the current user group Historical fit of channel C This is the sensitivity coefficient. For the current user group Channel fit threshold.

[0013] The server resource usage acquisition module includes: a server resource usage rate acquisition unit, used to acquire the CPU, memory, and bandwidth usage rates of the push server; and a server resource usage calculation unit, used to calculate the resource usage value of the push server according to a resource usage calculation formula; wherein the resource usage calculation formula is: In this formula, R represents the resource consumption of the push server. To measure the CPU utilization of the push server. To control the memory usage of the push server, To control the bandwidth utilization of the push server, W i The weights corresponding to each occupancy rate.

[0014] Thirdly, the present invention provides an intelligent message push device, including a memory and a processor. The memory is used to store computer program code and transmit the computer program code to the processor. The processor is used to execute the intelligent message push method described above according to the instructions in the computer program code.

[0015] In a fourth aspect, the present application provides a computer readable storage medium, wherein the computer readable storage medium stores computer executable instructions for causing a computer to execute the method as described above.

[0016] The present application has the following advantages: The present application acquires the server resource occupation condition through the quantitative index, and then adopts the message pushing strategy according to the server resource occupation value, adopts the first rule (i.e. the monomer customization strategy) to push the message when the server resource occupation value is less than the first threshold value, adopts the second rule (i.e. the group classification pushing strategy) to push the message when the server resource occupation value is greater than the first threshold value and less than the second threshold value, and adopts the default channel to push the message when the server resource occupation value is greater than the second threshold value. The present application overcomes the technical problems of a large amount of server resources occupied and a low success rate of pushing messages in the existing message pushing method, and realizes the optimal combination between the server resources and the pushing effect. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 is a structural schematic diagram of an embodiment of the existing message pushing method; Figure 2 is a structural schematic diagram of another embodiment of the existing message pushing method; Figure 3 is a structural schematic diagram of an embodiment of the intelligent message pushing method of the present application; Figure 4 is a flow schematic diagram of an embodiment of the intelligent message pushing method of the present application; Figure 5 is an embodiment of step S11 of Figure 4 ; Figure 6 is an embodiment of step S13 of Figure 4 ; Figure 7 is an embodiment of step S14 of Figure 4 ; Figure 8 is a structural schematic diagram of an embodiment of the intelligent message pushing device of the present application; Figure 9 is a structural schematic diagram of the server resource occupation acquisition module 11 of Figure 8 ; Figure 10 is a structural schematic diagram of the first message pushing module 13 of Figure 8 ; Figure 11 is a structural schematic diagram of the second message pushing module 14 of Figure 8 ; Figure 12 This is a schematic diagram of the structure of an embodiment of the intelligent message push device of the present invention. Detailed Implementation

[0018] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0019] Example 1

[0020] Please see Figure 4 , Figure 4 This is a flowchart illustrating an embodiment of the intelligent message push method of the present invention. Figure 4 As shown, the method includes the following steps: S11: Get the current resource usage value of the push server; Specifically, please refer to Figure 5 , Figure 5 yes Figure 4 A flowchart illustrating an embodiment of step S11. (See attached diagram.) Figure 5 As shown, step S11 includes the following steps: S111: Get the CPU, memory, and bandwidth utilization of the push server; CPU and memory usage are obtained using the `top` command in the operating system, while bandwidth usage is obtained using the `iftop` tool.

[0021] The CPU utilization, memory utilization, and bandwidth utilization values ​​are all within 0 to 1.

[0022] S112: Calculate the resource usage value of the push server according to the resource usage calculation formula; The formula for calculating resource usage is as follows: In this formula, R represents the resource consumption of the push server. To measure the CPU utilization of the push server. To control the memory usage of the push server, To control the bandwidth utilization of the push server, W i The weights are the corresponding to each occupancy rate.

[0023] S12: Compare the resource occupancy value with the preset first threshold and the preset second threshold; It should be noted that the second threshold is greater than the first threshold.

[0024] First, the resource usage value is compared with a preset first threshold. If the resource usage value is less than the preset first threshold, the comparison result is that the resource usage value is less than the preset first threshold. If the resource occupation value is greater than the preset first threshold value, the resource occupation value is compared with a preset second threshold value, and if the resource occupation value is less than the preset second threshold value, it is concluded that the resource occupation value is greater than the preset first threshold value and less than the preset second threshold value. If not, it is concluded that the resource occupation value is greater than the preset second threshold value.

[0025] In other embodiments, the resource occupation value can be compared with the preset second threshold value first, and then compared with the preset first threshold value.

[0026] S13: If the resource occupation value is less than the first threshold value, a message is pushed to the user based on a first rule; Specifically, please refer to Figure 6 , Figure 6 is Figure 4 a flowchart of an embodiment of step S13. As Figure 6 shown, step S13 includes the following steps: S131: Calculate the adaptation degree of each channel c for the current user u using a first adaptation degree calculation formula; The first adaptation degree calculation formula is as follows: In the formula, a is the adaptation degree of the current user u to the use of the channel c, is the historical reach rate of the current user u to the use of the channel c, is the activity of the reach target using the channel c, is the real-time occupation rate of the channel c, , , is the weight, , , The value range of a

[0027] The real-time occupation rate of the channel c refers to the quotient of the number of messages to be sent in the channel and the maximum number of messages that can be carried in the current channel queue. For example, an APP message sending channel supports a queue of 1000 messages, and the number of messages to be sent in the channel is 600. Therefore, the real-time occupation rate of the APP message sending channel is 60%.

[0028] The type of the channel includes: SMS, or APP, or WeChat, or email, which is not limited here. The following embodiments are only examples of SMS, APP, and WeChat channels.

[0029] According to the present step S131, the adaptation degree a of the WeChat channel for the current user u is calculated; According to the step S131, the adaptation degree b of the current user u to the short message channel is calculated. According to the step S131, the adaptation degree c of the current user u to the APP channel is calculated.

[0030] S132: The adaptation degrees of all channels are sorted according to the size, and the channel with the largest adaptation degree value is taken as the push channel of the current user u. The adaptation degrees a, b and c calculated in the step S131 are compared, if a > b > c, the channel (WeChat channel) corresponding to a is taken as the push channel of the current user u.

[0031] S14: If the resource occupation value is greater than the first threshold value and less than or equal to the second threshold value, a message is pushed to the user group based on a second rule; Specifically, please refer to Figure 7 , Figure 7 is Figure 4 the flowchart of an embodiment of the step S14. As shown in Figure 7 , the step S14 includes the following steps: S141: The second adaptation degree calculation formula is used to calculate the adaptation degree of the current user group to each channel c; The current user group is the user with the same label as the current user u.

[0032] The second adaptation degree calculation formula is as follows: In the formula, is the adaptation degree of the current user group to each channel c, is the number of the current user group, is the weight of the historical reach rate of the current user group to the use channel c, is the historical reach rate of the channel c, is the historical adaptation degree of the current user group to the channel c, is the sensitivity coefficient, is the channel adaptation degree threshold value of the current user group . Wherein, The calculation method of is the same as the method of calculating the first adaptation degree.

[0033] According to the step S141, the adaptation degree e of the current user group to the WeChat channel is calculated; According to the step S141, the adaptation degree f of the current user group to the short message channel is calculated; The current user group is calculated according to the step S141 The fitness g of the APP channel; S142: The fitness of all channels is sorted according to the size, and the channel with the largest fitness value is taken as the push channel of the current user group . The fitness e, f, g calculated in step S141 is compared, if e > f > g, the channel (WeChat channel) corresponding to e is taken as the push channel of the current user .

[0034] S15: If the resource occupation value is greater than the second threshold value, the default channel is used to push the message to the user.

[0035] The default channel is generally the SMS channel, which can be preset.

[0036] Actual scene example of the embodiment: Assume that the total number of push messages is 100,000, the total time is 30 minutes, the monitoring period is 5 minutes, the cpu, memory, and bandwidth weights set by the operator are 0.4, 0.3, and 0.3 respectively, the historical touch rate, touch target activity, and channel real-time occupation rate weights are set to 0.6, 0.3, and 0.1 respectively, and the first threshold value is 0.4 and the second threshold value is 0.8.

[0037] At T=0 minute, the cpu usage is 30%, the memory usage is 20%, and the bandwidth usage is 20%, so the resource occupation value of the push server at T=0 is: R=0.4×0.3+0.3×0.2+0.3×0.2=0.24.

[0038] Less than the first threshold value 0.4, adopt the first rule (i.e. Figure 3 The single customization strategy) push, during which the best channel is adapted to each user, for example, the WeChat, APP, and SMS channels are calculated for user A, the historical touch rates of the three channels are 90%, 85%, and 80% respectively, the real-time occupation rates are 20%, 30%, and 60% respectively, A opens WeChat for 9 days in the last 10 days, opens APP for 3 days, and clicks to view SMS for 4 days, and the above information is integrated to obtain: The fitness of the WeChat channel for the current user u . The fitness of the APP channel for the current user u . The fitness of the WeChat channel for the current user u .

[0039] Therefore, the WeChat channel is finally determined to be used for pushing to user A.

[0040] When T=5 minutes, the CPU utilization is 35%, the memory utilization is 30%, and the bandwidth utilization is 20%. Therefore, the resource usage of the push server at T=5 minutes is: R=0.4×0.35+0.3×0.3+0.3×0.2=0.29.

[0041] If it is still less than the first threshold of 0.4, continue to use the first rule (i.e. Figure 3 Push notifications using a customized individual strategy.

[0042] When T=10 minutes, the CPU utilization is 50%, the memory utilization is 40%, and the bandwidth utilization is 40%. Therefore, the resource usage of the push server at T=10 minutes is: R=0.4×0.5+0.3×0.4+0.3×0.4=0.44.

[0043] If R is greater than the first threshold of 0.4, the second rule is applied (i.e. Figure 3 (A segmented and categorized push strategy) is used to build user groups based on tags, adapt the best channels for them, and select users with active user tags to build G. active There are a total of 50,000 users. The historical success rate weighting is set to 0.7. Message push is a relatively mature technology that favors using historical data for decision-making. The sensitivity coefficient is the default 0.5. The current user group is G. active Historical fit of these three channels All values ​​are 0.5, and the channel compatibility threshold is set to 3. At this point, the historical push success rate for WeChat channels is 85%, for apps it is 95%, and for SMS it is 80%. For user group G... active Calculate the optimal channel: Calculate user group G active The compatibility with WeChat channels is as follows: ; Calculate user group G active The compatibility with the APP channel is as follows: ; Calculate user group G active The compatibility with SMS channels is as follows: .

[0044] Due to user group G active The app channel has the highest compatibility value, therefore, this group of users will send messages through the app messaging channel.

[0045] Continue monitoring R at T=15 and T=20. If R remains below the second threshold of 0.8, continue pushing using the second rule. When T=25, if R exceeds the second threshold, select the default method (SMS channel) for pushing.

[0046] It should be noted that various parameters of the embodiment, such as the first threshold value, the second threshold value, and various constants and parameters in the formula, are obtained by pre-configuration. All parameters are generated into a data dictionary, and an operator can quickly configure according to the message pushing effect.

[0047] Embodiment Two Please refer to Figure 8 , Figure 8 is a structural schematic diagram of an embodiment of the intelligent message pushing device of the application. As shown in Figure 8 , the intelligent message pushing device comprises a server resource occupation obtaining module 11, a resource occupation value comparison module 12, a first message pushing module 13, a second message pushing module 14, and a third message pushing module 15.

[0048] The server resource occupation obtaining module 11 is configured to obtain a resource occupation value of a pushing server at a current time. Specifically, please refer to Figure 9 , the server resource occupation obtaining module 11 comprises a server resource occupation rate obtaining unit 111 and a server resource occupation calculation unit 112. The server resource occupation rate obtaining unit 111 is configured to obtain the occupation rates of CPU, memory, and bandwidth of the pushing server. The server resource occupation calculation unit 112 is configured to calculate the resource occupation value of the pushing server according to a resource occupation calculation formula.

[0049] The resource occupation value comparison module 12 is configured to compare the resource occupation value with a preset first threshold value and a preset second threshold value.

[0050] The first message pushing module 13 is configured to push a message to a user based on a first rule if the resource occupation value is less than the first threshold value. Specifically, please refer to Figure 10 , the first message pushing module 13 comprises a first adaptation degree calculation unit 131 and a first channel determination unit 132. The first adaptation degree calculation unit 131 is configured to calculate the adaptation degree of each channel c for the current user u by using a first adaptation degree calculation formula. The first channel determination unit 132 is configured to sort the adaptation degrees of all channels according to the size, and take the channel with the largest adaptation degree value as the pushing channel of the current user u.

[0051] The second message pushing module 14 is configured to push a message to a user group based on a second rule if the resource occupation value is greater than the first threshold value and less than or equal to the second threshold value. Specifically, please refer to Figure 11 , the second message pushing module 14 comprises a second adaptation degree calculation unit 141 and a second channel determination unit 142. The second adaptation degree calculation unit 141 is configured to calculate the adaptation degree of each channel c for the current user group by using a second adaptation degree calculation formula. The adaptation degree of each channel c; the second channel determining unit is used for sorting the adaptation degrees of all channels according to the size, and taking the channel with the largest adaptation degree value as the push channel of the current user group .

[0052] The third message push module 15 adopts a default channel to push a message to the user if the resource occupation value is greater than the second threshold value.

[0053] Specifically, the working method of each module in the embodiment has been described in detail in Embodiment One, and will not be repeated here.

[0054] Embodiment Three Please refer to Figure 12 , Figure 12 is a structural schematic diagram of an embodiment of the intelligent message push device. As shown in the figure, Figure 12 the intelligent message push device comprises a memory and a processor, the memory is used for storing computer program code and transmitting the computer program code to the processor; the processor is used for executing the intelligent message push method as described in Embodiment One according to the instructions in the computer program code.

[0055] Embodiment Four The application further provides a computer readable storage medium, which stores computer executable instructions, and the computer executable instructions are used for enabling a computer to execute the method as described in Embodiment One.

[0056] The above is a specific description of the preferred implementation of the application, but the application is not limited to the described embodiments, and those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the application, and these equivalent modifications or replacements are all included in the range defined by the claims of the present application.

Claims

1. A smart message push method, characterized in that, include: Get the current resource usage of the push server; The resource occupancy value is compared with a preset first threshold and a preset second threshold. If the resource usage value is less than the first threshold, a message is pushed to the user based on the first rule; If the resource usage value is greater than the first threshold and less than or equal to the second threshold, then a message is pushed to the user group based on the second rule; If the resource usage value is greater than the second threshold, the default channel will be used to push messages to the user.

2. The intelligent message push method according to claim 1, characterized in that, The process of pushing messages to users based on the first rule includes: The adaptation degree of the current user u to each channel c is calculated using the first adaptation degree calculation formula; The adaptability of all channels is sorted by size, and the channel with the largest adaptability value is selected as the push channel for the current user u. The formula for calculating the first fit is as follows: In this formula, The fit between the current user u and the channel c. This represents the historical reach rate of the current user u to channel c. The activity level of reaching the target is determined by channel c. This represents the real-time occupancy rate of channel c. , , As weight, , , The value range is 0 to 1; The types of channels include: SMS, APP, WeChat, or email.

3. The intelligent message push method according to claim 2, characterized in that, The method of pushing messages to user groups based on the second rule includes: The current user group is calculated using the second fit calculation formula. The adaptation degree of each channel c, the current user group These are users with the same tags as the current user u; The fit scores of all channels are sorted by value, and the channel with the highest fit score is selected as the current user group. The push channel; The formula for calculating the second fit is as follows: In this formula, For the current user group The fit of each channel c, The number of users in the current user group. For the current user group The weighting of historical reach rates using channel c This represents the historical reach rate of channel C. For the current user group Historical fit of channel C This is the sensitivity coefficient. For the current user group Channel fit threshold.

4. The intelligent message push method according to claim 1, characterized in that, The step of obtaining the current resource usage value of the push server includes: Obtain the CPU, memory, and bandwidth utilization of the push server; The resource usage value of the push server is calculated according to the resource usage calculation formula; The formula for calculating resource usage is as follows: In this formula, R represents the resource consumption of the push server. To measure the CPU utilization of the push server. To control the memory usage of the push server, To control the bandwidth utilization of the push server, W i The weights corresponding to each occupancy rate.

5. An intelligent message push device, characterized in that, include: The server resource usage acquisition module is used to obtain the current resource usage value of the push server. The resource occupancy comparison module is used to compare the resource occupancy value with a preset first threshold and a preset second threshold; The first message push module is used to push a message to the user based on a first rule if the resource usage value is less than the first threshold. The second message push module is used to push messages to the user group based on the second rule if the resource occupancy value is greater than the first threshold and less than or equal to the second threshold. The third message push module is used to push messages to the user using a default channel if the resource usage value is greater than the second threshold.

6. The intelligent message push device according to claim 5, characterized in that, The first message push module includes: The first adaptation calculation unit is used to calculate the adaptation degree of the current user u to each channel c using the first adaptation calculation formula. The first channel determination unit is used to sort all channels by their adaptability and select the channel with the highest adaptability value as the push channel for the current user u. The formula for calculating the first fit is as follows: In this formula, The fit between the current user u and the channel c. This represents the historical reach rate of the current user u to channel c. The activity level of reaching the target is determined by channel c. This represents the real-time occupancy rate of channel c. , , As weight, , , The value range is 0 to 1.

7. The intelligent message push device according to claim 5, characterized in that, The second message push module includes: The second fit calculation unit is used to calculate the current user group using the second fit calculation formula. The degree of adaptation to each channel c; The second channel determination unit is used to sort all channels by their fit scores and select the channel with the highest fit score as the current user group. The push channel; The formula for calculating the second fit is as follows: In this formula, For the current user group The fit of each channel c, The number of users in the current user group. For the current user group The weighting of historical reach rates using channel c This represents the historical reach rate of channel C. For the current user group Historical fit of channel C This is the sensitivity coefficient. For the current user group Channel fit threshold.

8. The intelligent message push device according to claim 5, characterized in that, The server resource usage acquisition module includes: The server resource utilization rate acquisition unit is used to acquire the CPU, memory, and bandwidth utilization rates of the push server. The server resource usage calculation unit is used to calculate the resource usage value of the push server according to the resource usage calculation formula. The formula for calculating resource usage is as follows: In this formula, R represents the resource consumption of the push server. To measure the CPU utilization of the push server. To control the memory usage of the push server, To control the bandwidth utilization of the push server, W i The weights corresponding to each occupancy rate.

9. A smart message push device, characterized in that, The device includes a memory and a processor. The memory is used to store computer program code and transmit the computer program code to the processor. The processor is used to execute the intelligent message push method as described in any one of claims 1 to 4 according to the instructions in the computer program code.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions for causing a computer to perform the method as described in any one of claims 1 to 4.

Images