WO2018014846A1 - 一种应用消息推送方法、装置 - Google Patents

一种应用消息推送方法、装置 Download PDF

Info

Publication number
WO2018014846A1
WO2018014846A1 PCT/CN2017/093534 CN2017093534W WO2018014846A1 WO 2018014846 A1 WO2018014846 A1 WO 2018014846A1 CN 2017093534 W CN2017093534 W CN 2017093534W WO 2018014846 A1 WO2018014846 A1 WO 2018014846A1
Authority
WO
WIPO (PCT)
Prior art keywords
user equipment
message
response rate
time period
sent
Prior art date
Application number
PCT/CN2017/093534
Other languages
English (en)
French (fr)
Inventor
岳赟
陶宇田
钟明洁
曹志远
吴文汇
白海波
Original Assignee
阿里巴巴集团控股有限公司
岳赟
陶宇田
钟明洁
曹志远
吴文汇
白海波
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 阿里巴巴集团控股有限公司, 岳赟, 陶宇田, 钟明洁, 曹志远, 吴文汇, 白海波 filed Critical 阿里巴巴集团控股有限公司
Priority to SG11201900503VA priority Critical patent/SG11201900503VA/en
Priority to EP17830480.4A priority patent/EP3490198B1/en
Priority to PL17830480T priority patent/PL3490198T3/pl
Priority to JP2019503319A priority patent/JP6765500B2/ja
Priority to KR1020197005197A priority patent/KR102192877B1/ko
Priority to ES17830480T priority patent/ES2818610T3/es
Priority to EP20189068.8A priority patent/EP3751797B1/en
Publication of WO2018014846A1 publication Critical patent/WO2018014846A1/zh
Priority to US16/254,329 priority patent/US10812607B2/en
Priority to US16/723,676 priority patent/US10958749B2/en

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L51/00—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
    • H04L51/21—Monitoring or handling of messages
    • H04L51/224—Monitoring or handling of messages providing notification on incoming messages, e.g. pushed notifications of received messages
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L51/00—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
    • H04L51/21—Monitoring or handling of messages
    • H04L51/214—Monitoring or handling of messages using selective forwarding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/955—Retrieval from the web using information identifiers, e.g. uniform resource locators [URL]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/955—Retrieval from the web using information identifiers, e.g. uniform resource locators [URL]
    • G06F16/9562—Bookmark management
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L51/00—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail
    • H04L51/07—User-to-user messaging in packet-switching networks, transmitted according to store-and-forward or real-time protocols, e.g. e-mail characterised by the inclusion of specific contents
    • H04L51/18—Commands or executable codes
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/14—Session management
    • H04L67/146—Markers for unambiguous identification of a particular session, e.g. session cookie or URL-encoding
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/55—Push-based network services

Definitions

  • the present application belongs to the field of information processing technologies, and in particular, to an application message pushing method and apparatus.
  • smartphones have the push function of notification messages.
  • various types of applications installed in smart phones often send relevant marketing or service messages to users through mobile phones, such as promotion of products.
  • the existing bulk message push method basically selects the content to be pushed, and the background configuration personnel specify a sending time for batch sending, and all users receive the pushed content at almost the same time.
  • the existing batch message pushing method does not consider whether the user will click on the received message to read, and assume that the user does not click on the message when it is working or in the meeting when receiving the message. After a long time, the user may have forgotten the received messages, and then will not click to read the messages, then, the purpose of various applications to push the message in order to improve their open rate can not be achieved.
  • the present application provides an application message pushing method and apparatus, which can solve the problem that the existing message delivery method has a low response rate in a timely manner.
  • an application message pushing method including:
  • the tag of the message to be sent is detected as a message tag that is not sent in real time
  • determining, according to the message timely response rate and the correction policy, the optimal time period for each user equipment response message in the user equipment group includes:
  • the time interval corresponding to the maximum value of the corrected message timely response rate is determined as the optimal time period.
  • the method further includes:
  • the method further includes:
  • the method further includes:
  • each user equipment in the user equipment group is re-determined in the optimal time period.
  • the method further includes:
  • the optimal time period of the user equipment response message is re-determined.
  • the method further includes:
  • the tag of the message to be sent is detected as a message tag sent in real time
  • the sending the to-be-sent message to the corresponding user equipment in the user equipment group includes:
  • the second aspect of the present application provides an application message pushing apparatus, including:
  • a detecting module configured to detect that the label of the message to be sent is a message label that is not sent in real time
  • a determining module configured to dynamically determine an optimal time period of each user equipment response message in the user equipment group according to the message timely response rate and the correction policy for the user equipment group of the message to be sent;
  • the sending module is configured to send the to-be-sent message to the corresponding user equipment in the user equipment group according to an optimal time period of each user equipment response message.
  • the device further includes:
  • a preset module configured to preset a time interval of a timely response rate of the statistical message
  • a statistics module configured to count, according to a preset time interval of the preset module, a quantity of messages received by each user equipment in each time interval and a number of messages that are timely responded;
  • the determining module is configured to determine, according to the number of messages received by each user equipment in each time interval and the number of messages that are timely responded, according to the statistics module, that each user equipment has timely messages in each time interval. Response rate
  • the determining module is configured to determine a time interval corresponding to a maximum value of the timely response rate of the message as the optimal time period;
  • a correction module configured to correct, by using a preset correction policy, a timely response rate of each user equipment determined by the determining module in each time interval, so as to minimize a fluctuation range of the timely response rate of the message;
  • the determining module is configured to determine, as the optimal time period, a time interval corresponding to a maximum value of the timely response rate of the corrected message.
  • the determining module is further configured to:
  • the device further includes:
  • a grouping module configured to randomly divide the user equipment group into a first user equipment group and a second user equipment group according to a preset ratio
  • the sending module is further configured to send the to-be-sent message to the first user equipment according to an optimal time period of each user equipment response message in the first user equipment group determined by the determining module The corresponding user equipment in the group;
  • the sending module is further configured to randomly send the to-be-sent message to each user equipment in the second user equipment group.
  • the statistics module is further configured to: collect a timely response rate of each user equipment in the first user equipment group in the optimal time period, and each user in the second user equipment group.
  • the timely response rate of the device's message is further configured to: collect a timely response rate of each user equipment in the first user equipment group in the optimal time period, and each user in the second user equipment group. The timely response rate of the device's message;
  • the device also includes:
  • a comparison module configured to compare, when the user equipment in the first user equipment group is in the optimal time period, a message timely response rate and a message of each user equipment in the second user equipment group If the difference between the timely response rates is less than or equal to the preset gap threshold, then the determining module determines that each user equipment in the user equipment group is in the optimal time period.
  • the determining module is further configured to: when determining that the timely response rate of the user equipment in the optimal time period is lower than a preset message timely response rate threshold in the user equipment group, The optimal time period for the user equipment response message.
  • the detecting module is further configured to detect that the label of the message to be sent is sent in real time. Message label
  • the sending module is further configured to send the to-be-sent message to each user equipment in the user equipment group in real time.
  • the application further provides an application backend server, comprising: the application message pushing device described in the second aspect above.
  • the optimal time period of each user equipment response message in the user equipment group is determined according to the user equipment group of the message to be sent, that is, the maximum probability time period of each user click message may be obtained, according to each The user equipment response message is sent to the corresponding user equipment in the user equipment group, so that different users receive the time at different time points.
  • Push messages to the system to maximize the probability of users clicking on the message Since the present invention fully considers the point-in-time preference of the user to click on the message, the user customizes the time point at which the user receives the message, and pushes the message through the personalized time point, thereby maximizing the probability of the user clicking the message, thereby greatly improving the effectiveness of the message hitting the target user. Sex. Therefore, the problem that the timely response rate of the message is low in the batch message pushing method existing in the prior art can be solved.
  • the invention can also take into account that the time point preference of the user clicks on the message changes, and adaptively corrects the optimal time period of the user equipment response message, thereby ensuring the validity of the message hitting the target user.
  • FIG. 1 is a schematic flowchart of an application message pushing method according to an embodiment of the present application
  • FIG. 2 is a schematic flow chart of an implementation method of step 102 in the first embodiment shown in FIG. 1;
  • FIG. 3 is a schematic flow chart of still another implementation method of step 102 in the first embodiment shown in FIG. 1;
  • FIG. 4 is a schematic flowchart of an application message pushing method according to an embodiment of the present application.
  • FIG. 5 is a schematic diagram of a specific implementation function of an application message pushing method according to an embodiment of the present application
  • FIG. 6 is a schematic structural diagram of an application message pushing apparatus according to an embodiment of the present disclosure.
  • FIG. 7 is a schematic structural diagram of an application backend server according to an embodiment of the present disclosure.
  • the day is divided into several time segments, and the number of browsing times in different time segments is compared according to the number of times the user browses the message in each time period;
  • the time period with the most browsing times is set to the time period with the highest priority. Further, it is determined whether the time when the terminal receives the message falls within the highest priority time period, and if so, the receiving message is sent to the user in real time, otherwise the push message is postponed to user.
  • the inventors found in the process of implementing the present invention that the following problems exist in the prior art:
  • the time period with the highest priority when setting the time period with the highest priority, only the time period of the user's maximum number of browsing times is taken as the optimal time period, but the premise of the user browsing depends on whether the message has been received or not. It is very likely that the user is now idle. However, the terminal does not have a message, and the user does not browse. Therefore, in the prior art, the setting of the highest priority time period does not accurately reflect the preference of the user to click on the message. Moreover, in the prior art, once the time period with the highest priority is set, the user only receives the received message pushed by the terminal during the time period with the highest priority. When the user is accustomed to change, the adaptive adjustment cannot be performed with the highest priority. Time period.
  • a terminal such as a mobile phone receives a message generated by each application background or an operating system, and then pushes the received message to the user according to the highest priority time period set above. Therefore, the push message is performed by the terminal.
  • the terminal since all the message generation is not the operating system itself, and the message generated by each application in the background is included, it is difficult for the terminal to distinguish the message with high real-time requirement and low real-time requirement, such as Chat messages, breaking news, consumer account reminders and other real-time requirements are very high, and the terminal needs to be pushed to the user immediately.
  • the highest priority time period is set, the user will only be in this The highest priority period is received by the terminal. Receiving messages, such that messages with high real-time requirements cannot be pushed to the user in real time, causing users to miss important real-time messages.
  • the setting of the highest priority time period lacks an adaptive adjustment mechanism (compared to mechanical stereotype), and does not accurately reflect the preference of the user to click on the message, so that the effectiveness of the message hitting the target user cannot be effectively improved. Second, it will cause users to miss important real-time messages.
  • the present invention personally adjusts the time point preference of each user to click a message through an adaptive adjustment mechanism, thereby realizing a personalized user to receive a message at a point in time, through a personalized user.
  • the time point of receiving the message pushes the message to the user, thereby maximizing the probability that the user clicks on the message, and improving the effectiveness of the message hitting the target user.
  • FIG. 1 is a schematic flowchart of an application message pushing method according to an embodiment of the present application; as shown in FIG. 1, the method includes:
  • each message carries a message tag, where the message tag includes a message tag that is not sent in real time and a message tag that is sent in real time. For this reason, according to the message label carried in the message, when the message label is a non-real-time sent message label, it can be determined that the message to be sent is a non-real-time sent message. When the message label is a real-time sent message label, it can be determined that the message to be sent is A message sent in real time.
  • step 102 When it is detected that the message to be sent is a message that is not sent in real time, step 102 is performed. When it is detected that the message to be sent is a message sent in real time, step 104 is performed.
  • the user equipment group of the to-be-sent message dynamically determines an optimal time period of each user equipment response message in the user equipment group according to a message timely response rate and a modification policy.
  • the user equipment in the embodiment of the present invention may be, for example, a smart phone, a PAD (tablet computer), a PC tablet, and the like.
  • the user equipment may have a wired network interface or a wireless network interface.
  • an application background such as a mobile phone Taobao (an application APP in a mobile phone) needs to send a product promotion message to a user mobile phone in batches. Since the number of mobile phones installed with mobile phone Taobao is large, the user mobile phone here belongs to a user equipment. Therefore, it is called user setting in the embodiment of the present invention. Prepare the group.
  • the embodiment of the present invention needs to acquire the time point preference of each user click message in the user group. It should be noted that, when the user equipment receives the product promotion message sent by the application of the mobile phone Taobao, the user may also refer to the product promotion message by clicking the product promotion message through the user equipment. Therefore, the step 101 determines that the product promotion message is The optimal time period for each user equipment response message in the user equipment group is essentially the time point preference for obtaining each user click message in the user group.
  • the current mobile phone bulk message push (excluding the message triggered by the event rule in real time) starts at a specified time point T, and the real touch reaches the user only depends on the time point T and the time of the system push processing. Since the probability that the user clicks on the message is exponentially attenuated as the time of receiving the message, that is, most users will only click on the message within a short period of time (usually within half an hour), so whether the user clicks on the message and the user receives the message. The point in time is strongly correlated.
  • the invention considers the time point preference of the user to receive the message, and can send the batch push message to different users at different time points, that is, personally adjust the time point preference of each user to click the message, thereby realizing the customized user receiving the message. At the time point, the user pushes the message at the time point when the personalized user receives the message, thereby maximizing the probability of the user clicking the message and improving the effectiveness of the message hitting the target user.
  • each mobile phone can be analyzed, for example, by counting the history (for example, the last year) of each mobile phone Taobao user (equivalent to the user equipment installed with the mobile phone Taobao). The maximum probability period for the Taobao user to click on the message (equivalent to the best time period for the response message).
  • Step 102 may be referred to the detailed description in the embodiment shown in FIG. 2 and FIG. 3 during specific implementation.
  • the maximum probability time period of each mobile phone Taobao user click message (equivalent to the best time period of the response message), the mobile phone Taobao application background will be the commodity maximum promotion time period (the optimal time period) ) are sent to the corresponding mobile phone Taobao users.
  • the application background of the mobile phone Taobao is The product promotion message is sent to the user equipment used by the user A between 12 noon and 12:10, so that the user A clicks the message very quickly when receiving the message. In this way, the effectiveness of the message hitting the target user is improved.
  • the optimal time period of each user equipment response message in the user equipment group is determined according to the user equipment group of the message to be sent, that is, the maximum probability time period of each user click message may be obtained, according to each The user equipment response message is sent to the corresponding user equipment in the user equipment group, so that different users receive the time at different time points.
  • Push messages to the system to maximize the probability of users clicking on the message Since the present invention fully considers the point-in-time preference of the user to click on the message, the user customizes the time point at which the user receives the message, and pushes the message through the personalized time point, thereby maximizing the probability of the user clicking the message, thereby greatly improving the effectiveness of the message hitting the target user. Sex. Therefore, the problem that the timely response rate of the message is low in the batch message pushing method existing in the prior art can be solved.
  • step 102 each user equipment response message in the user equipment group is determined.
  • the optimal time period may specifically include the following:
  • Count according to a preset time interval, a quantity of messages received by each user equipment in each time interval and a number of messages that respond in time;
  • This step is mainly to count the situation in which the user history receives the message and clicks the message.
  • the statistical sample that is, the case where the user history receives the message and clicks the message
  • the statistics of the user receiving the message and the click message in the last year can be counted.
  • the number of messages received by each user equipment in each time interval and the number of messages responding in time are counted. .
  • the working days and holidays can be distinguished in the statistics to perform corresponding differential statistics.
  • the number of timely response messages of each user equipment in each time interval is divided by the received number.
  • the number of messages can obtain the timely response rate of the user equipment during the time interval, so that the timely response rate of each user equipment in each time interval can be obtained.
  • the timely response rate of the message in a certain time interval indicates that the user clicks the message in time when receiving the message.
  • the probability is high. Therefore, the time interval corresponding to the maximum response rate of the message can be determined as the optimal time period of the user equipment response message.
  • FIG. 3 is a schematic flowchart of still another implementation method of step 102 in the first embodiment shown in FIG. 1.
  • the number of messages received by the user equipment in each time interval may be different. Larger, causing a large fluctuation of the response rate of the message in each time interval, thereby affecting the optimal time period for accurately determining the response message of each user equipment. For this reason, each time interval is required in the embodiment of the present invention.
  • the timely response rate of the message is corrected; specifically, step 102 may include the following:
  • Count according to a preset time interval, a quantity of messages received by each user equipment in each time interval and a number of messages that respond in time;
  • This step is mainly to count the situation in which the user history receives the message and clicks the message.
  • the statistical sample that is, the case where the user history receives the message and clicks the message
  • each user device is counted in the last year, at each time. The number of messages received during the interval and the number of messages that responded in a timely manner.
  • the working days and holidays can be distinguished in the statistics to perform corresponding differential statistics.
  • the number of timely response messages of each user equipment in each time interval is divided by the received number.
  • the number of messages can obtain the timely response rate of the user equipment during the time interval, so that the timely response rate of each user equipment in each time interval can be obtained.
  • the preset correction policy may be modified, for example, by using a Wilson interval.
  • the user receives 10 messages at time interval 1, clicks 6 messages, receives 100 messages at time interval 2, and clicks 60 messages.
  • the calculated click probability is 0.6.
  • the user's click probabilities at interval 1 and interval 2 are 0.31 and 0.50, respectively (the lower limit of the normal distribution interval with a confidence of 0.95), that is, the time is considered.
  • the click probability calculated by the interval 2 by frequency is more stable and the fluctuation range is smaller.
  • the optimal time period of the user equipment response message may be determined more accurately. For example, the time interval corresponding to the maximum response rate of the plurality of messages is obtained, and the time interval of the earliest time is determined as the optimal time period of the user equipment response message; for example, the time interval in which the network state is optimal may be determined as The optimal time period for the user device to respond to the message.
  • the user equipment group is randomly grouped according to a preset ratio, for a group of user equipment groups, in order to ensure that the time point preference of the user clicks on the message changes.
  • the method of randomly selecting the time point to push the message is adopted, and then the effects of the two sets of user equipment response messages are compared, according to the effect ratio.
  • the feedback information of the message pushing time of the user equipment or the two groups of users is used to determine whether to recalculate the optimal time period of the user equipment response message, and the specific implementation includes the following:
  • the user equipment group is randomly divided into a first user equipment group and a second user equipment group according to a preset ratio.
  • a user A may be assigned to the first user equipment group or may be assigned to the second user equipment group each time the user pushes;
  • each user equipment in the user equipment group is re-determined in the optimal time period.
  • the gap threshold of the timely response rate of the message may be preset, and the response rate and the timely response rate of each user equipment in the first user equipment group during the optimal time period are obtained.
  • the difference between the message and the response rate of each user equipment in the second user equipment group is less than or equal to a preset gap threshold, it indicates that there is a problem in the optimal time period of the user equipment response message obtained by the foregoing statistics, and needs to be determined again. .
  • a message timely response rate threshold may be preset, and when it is determined that the user equipment in the user equipment group has a timely response rate of the message in the optimal time period is lower than a preset message.
  • Responsibility rate threshold need to re-determine the best response of the user equipment response message
  • the time period indicates that the time point preference of the user equipment of the user equipment may change.
  • the embodiment of the present invention can adjust the optimal time period of the user equipment response message in time. To ensure the accuracy of the best time period for the user device to respond to the message.
  • the user equipment group is randomly grouped according to a preset ratio, and for one group of user equipment groups, the message is pushed in the best time period of the response message, and the other group of user equipment groups is randomly used.
  • Select the method of pushing the message at the time point and then compare the effects of the two sets of user equipment response messages, and determine whether the user equipment response message is optimal according to the effect comparison or the feedback information of the two groups of user equipments on the message pushing time.
  • the time period is recalculated, and the time point preference of the user clicking the message may be considered, and the optimal time period of the user equipment response message is adaptively corrected, thereby ensuring the validity of the message hitting the target user.
  • FIG. 5 is a schematic diagram of a specific implementation function of an application message pushing method according to an embodiment of the present application. As shown in FIG. 5, the method includes:
  • Step 1 User time point preference calculation engine
  • the maximum probability interval of the user clicking the message is calculated by receiving the message and clicking the message in the user history (the last one year), as follows:
  • the Wilson interval is used for correction.
  • the user receives 10 messages at time interval 1, clicks 6 messages, and receives messages at time interval 2.
  • 100 click on 60
  • the calculated click probability is 0.6
  • the user's click probability at time interval 1 and interval 2 is 0.31 and 0.50 respectively (the confidence interval is 0.95 for the normal distribution interval)
  • the lower limit that is, the click probability calculated by the time interval 2 by the frequency is more stable and the fluctuation range is smaller.
  • Step 2 Select the content and user group to send
  • the content can be specified by the business personnel or calculated by the system.
  • the user group can be a full user or a user filtered by a tag, such as a user aged 18-22, a female user, and the like.
  • Step 3 Sending a message
  • the sent content and the user group are pushed to the message sending device, and after the user selects the best push time point, the user sends the message to the user at the optimal push time point, which specifically includes:
  • User group grouping means The background asynchronous thread combines the user groups specified in step 2 with the user time point preference information calculated in step 1, and groups the users with the time interval ⁇ t mentioned in step 1. For users who do not have explicit point-in-time preference information, they are randomly assigned to a transmission time packet.
  • bucket device after the user group is finished, through the bucket device, a certain proportion of users (such as 20%) transmission time is set to random transmission time, which is used for effect comparison and collects more user-to-message sending time. Feedback information.
  • User message sending device the user message sending device will start the task executing thread at a time interval of ⁇ t on multiple cluster machines through distributed concurrent scheduling control, and concurrently query (2) the bucket device in the system During the time period, the user sends data and sends a trigger message push.
  • the user message sending device will start the task executing thread at a time interval of ⁇ t on multiple cluster machines through distributed concurrent scheduling control, and concurrently query (2) the bucket device in the system During the time period, the user sends data and sends a trigger message push.
  • the distributed system for example: A, B, C three machines have received the sending task for user 1 at the same time point, if not processed, A, B, C
  • the message push for user 1 will be triggered separately.
  • the system will directly perform the logical grouping according to the user ID dimension and the execution task machine to ensure that only one machine of one user is responsible for message push.
  • the identifier of the sending server corresponding to the identifier of each user equipment is determined according to the identifier of each user equipment in the user equipment group, so that the sending server sends the to-be-sent message Send to the corresponding user equipment; the specific practices are as follows:
  • Log the task execution machine logically with a starting number of 0. For example, if there are 20 machines performing tasks, then the logical number is all integers from 0-19.
  • b Mapping the user's ID information to the machine number. Taking the above example as an example, assuming that the ID of the user 1 is 123, then the pair 123 is modulo 3 for 20, which means that the user 1 needs to be responsible for message push by the machine with logical number 3.
  • the embodiment of the present invention fully considers the time point preference of the user to click on the message, personalizes the time point at which the user receives the message, and pushes the message through the personalized time point, thereby maximizing the probability of the user clicking the message. At the same time, it also takes into account that the time point preference of the user clicks on the message changes, and adaptively corrects the optimal time period of the user equipment response message, thereby ensuring the validity of the message hitting the target user.
  • FIG. 6 is a schematic structural diagram of an application message pushing apparatus according to an embodiment of the present disclosure. As shown in FIG. 6, the method includes:
  • the detecting module 60 is configured to detect that the label of the message to be sent is a message label that is not sent in real time;
  • the determining module 61 is configured to dynamically determine an optimal time period of each user equipment response message in the user equipment group according to the message timely response rate and the correction policy for the user equipment group of the message to be sent;
  • the sending module 62 is configured to send the to-be-sent message to the corresponding user equipment in the user equipment group according to an optimal time period of each user equipment response message.
  • the device further comprises:
  • the preset module 63 is configured to preset a time interval of a timely response rate of the statistical message
  • the statistics module 64 is configured to count, according to the preset time interval of the preset module, the number of messages received by each user equipment in each time interval and the number of messages that are timely responded;
  • the determining module 61 is configured to determine, as the optimal time period, a time interval corresponding to a maximum value of the timely response rate of the message.
  • the device further includes:
  • the correction module 65 is configured to correct, by using a preset correction policy, a timely response rate of each user equipment determined by the determining module in each time interval, so as to minimize the fluctuation of the timely response rate of the message. ;
  • the determining module 61 is configured to determine, as the optimal time period, a time interval corresponding to a maximum value of the timely response rate of the corrected message.
  • the determining module 61 is further configured to: if the maximum value of the message timely response rate of the user equipment includes multiple:
  • the device further includes:
  • the grouping module 66 is configured to randomly divide the user equipment group into a first user equipment group and a second user equipment group according to a preset ratio;
  • the sending module 62 is further configured to send the to-be-sent message to the first user separately according to an optimal time period of each user equipment response message in the first user equipment group determined by the determining module The corresponding user equipment in the equipment group;
  • the sending module 62 is further configured to randomly send the to-be-sent message to each user equipment in the second user equipment group.
  • the statistic module 64 is further configured to collect a message timely response rate of each user equipment in the first user equipment group in the optimal time period, and each user equipment in the second user equipment group. The timely response rate of the message.
  • the device also includes:
  • the comparison module 67 is configured to compare the timely response rate of each user equipment in the first user equipment group during the optimal time period with each user equipment in the second user equipment group. If the difference between the timely response rates of the messages is less than or equal to the preset gap threshold, then the determining module 61 determines that each user equipment in the user equipment group is in the optimal time period.
  • the determining module 61 is further configured to: when determining that the timely response rate of the message of the user equipment in the optimal time period is lower than a preset message timely response rate threshold in the user equipment group, The optimal time period for the user device to respond to the message.
  • the detecting module 60 is further configured to detect that the label of the message to be sent is a message label sent in real time;
  • the sending module 62 is further configured to send the to-be-sent message to each user equipment in the user equipment group in real time.
  • the device of the embodiment of the present invention fully considers the time point preference of the user to click on the message, personalizes the time point at which the user receives the message, and pushes the message through the personalized time point, thereby maximizing the probability of the user clicking the message. At the same time, it also takes into account that the time point preference of the user clicks on the message changes, and adaptively corrects the optimal time period of the user equipment response message, thereby ensuring the validity of the message hitting the target user.
  • FIG. 7 is a schematic structural diagram of an application backend server according to an embodiment of the present disclosure, where the background server may have a wired network interface or a wireless network interface. As shown in FIG. 7, comprising: a memory 10; one or more processors 11;
  • the memory 10 stores instructions for implementing the message pushing method described in any of the above embodiments in FIG. 1 to FIG. 5.
  • the processor 11 calls the instruction saved in the memory 10, the following steps may be performed:
  • the tag of the message to be sent is detected as a message tag that is not sent in real time
  • the determining the optimal time period for each user equipment response message in the user equipment group includes:
  • the time interval corresponding to the maximum value of the timely response rate of the message is determined as the optimal time period.
  • the method After determining the timely response rate of each user equipment in each time interval, the method includes:
  • the time interval corresponding to the maximum value of the corrected message timely response rate is determined as the optimal time period.
  • the processor 11 may further perform the following steps:
  • processor 11 can also perform the following steps:
  • processor 11 can also perform the following steps:
  • the optimal time period of the user equipment response message is re-determined.
  • the sending the to-be-sent message to the corresponding user equipment in the user equipment group includes:
  • processor 11 can also perform the following steps:
  • the tag of the message to be sent is detected as a message tag sent in real time
  • the background service device in the embodiment of the present invention fully considers the time point preference of the user to click on the message, personalizes the time point at which the user receives the message, and pushes the message through the personalized time point, thereby maximizing the probability of the user clicking the message. At the same time, it also takes into account that the time point preference of the user clicks on the message changes, and adaptively corrects the optimal time period of the user equipment response message, thereby ensuring the validity of the message hitting the target user.
  • a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
  • processors CPUs
  • input/output interfaces network interfaces
  • memory volatile and non-volatile memory
  • the memory may include non-persistent memory, random access memory (RAM) or non-volatile memory, such as read only memory (ROM) or flash memory (flash RAM), in a computer readable medium.
  • RAM random access memory
  • ROM read only memory
  • flash RAM flash memory
  • Memory is an example of a computer readable medium.
  • Computer readable media includes both permanent and non-persistent, removable and non-removable media.
  • Information storage can be implemented by any method or technology.
  • the information can be computer readable instructions, data structures, modules of programs, or other data.
  • Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory. (ROM), electrically erasable programmable read only memory (EEPROM), flash memory or other memory technology, compact disk read only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, Magnetic tape cartridges, magnetic tape storage or other magnetic storage devices or any other non-transportable media can be used to store information that can be accessed by a computing device.
  • computer readable media does not include non-transitory computer readable media, such as modulated data signals and carrier waves.
  • first device if a first device is coupled to a second device, the first device can be directly electrically coupled to the second device, or by other devices or coupling means.
  • the grounding is electrically coupled to the second device.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Transfer Between Computers (AREA)
  • Telephonic Communication Services (AREA)
  • User Interface Of Digital Computer (AREA)

Abstract

本申请公开了一种应用消息推送方法、装置，所述方法包括根据待发送消息的用户设备群，确定所述用户设备群中每个用户设备响应消息的最佳时间段；根据每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述用户设备群中对应的用户设备，可以解决现有的批量消息推送方法中存在消息及时响应率较低的问题。

Description

一种应用消息推送方法、装置 技术领域
本申请属于信息处理技术领域，具体地说，涉及一种应用消息推送方法、装置。
背景技术
目前智能手机均具有通知消息的推送功能，安装在智能手机中的各类应用为了提高自身的打开率(点击率)，往往会通过手机向用户批量发送相关的营销或者服务消息，例如商品的促销、热门视频、资讯等等消息，现有的批量消息推送方法基本是选定需要推送的内容，由后台配置人员指定一个发送时间进行批量发送，所有用户几乎在同一时间收到被推送的内容。
然而，现有的批量消息推送方法中没有考虑到用户是否会点击这些接收到的消息进行阅读，假设用户接收到消息的时候正在工作中或会议中，在短时间内不会去点击这些消息，那么时间一长，用户可能已经忘记这些接收到的消息，于是也不会点击阅读这些消息了，那么，各类应用为了提高自身的打开率而推送消息的目的就达不到了。
因此，现有的批量消息推送方法中存在消息及时响应率较低的问题，从而带来消息命中目标用户的有效性下降。
发明内容
有鉴于此，本申请提供一种应用消息推送方法、装置，可以解决现有的批量消息推送方法中存在消息及时响应率较低的问题。
为了解决上述技术问题，本申请第一方面提供一种应用消息推送方法，包括：
检测到待发送消息的标签为非实时发送的消息标签；
针对所述待发送消息的用户设备群，根据消息及时响应率以及修正策略动态确定所述用户设备群中每个用户设备响应消息的最佳时间段；
根据每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述用户设备群中对应的用户设备。
可选地，根据消息及时响应率以及修正策略动态确定所述用户设备群中每个用户设备响应消息的最佳时间段包括：
预设统计消息及时响应率的时间间隔；
根据预设的时间间隔，统计每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量；
根据每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量，确定每个用户设备在每个时间间隔内的消息及时响应率；
将所述消息及时响应率最大值对应的时间间隔确定为所述最佳时间段；
利用预设的修正策略对所述确定的每个用户设备在每个时间间隔内的消息及时响应率进行修正，以使所述消息及时响应率的波动幅度最小化；
对应地，将所述修正后的消息及时响应率的最大值对应的时间间隔确定为所述最佳时间段。
可选地，若所述用户设备的消息及时响应率最大值包括多个时，所述方法还包括：
分别获取多个所述消息及时响应率最大值对应的时间间隔，将时间最早的时间间隔确定为所述用户设备响应消息的最佳时间段；或者将网络状态最优的时间间隔确定为所述用户设备响应消息的最佳时间段。
可选地，所述的方法还包括：
根据预设的比例，将所述用户设备群分为第一用户设备群和第二用户设备群；
根据所述第一用户设备群中每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述第一用户设备群中对应的用户设备；
将所述待发送消息随机发送给所述第二用户设备群中的每个用户设备。
可选地，所述的方法还包括：
统计所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率；
当比对得到所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应 率之间的差距小于等于预设的差距阈值，则重新确定所述用户设备群中每个用户设备在所述最佳时间段。
可选地，所述的方法还包括：
当确定所述用户设备群中存在用户设备在所述最佳时间段的消息及时响应率低于预设的消息及时响应率阈值时，重新确定所述用户设备响应消息的最佳时间段。
可选地，所述的方法还包括：
检测到待发送消息的标签为实时发送的消息标签；
将所述待发送消息实时发送给所述用户设备群中每个用户设备。
可选地，将所述待发送消息分别发送给所述用户设备群中对应的用户设备包括：
根据所述用户设备群中每个用户设备的标识，确定与每个用户设备的标识对应的发送服务器的标识，以使所述发送服务器将所述待发送消息发送给对应的用户设备。
本申请第二方面提供一种应用消息推送装置，包括：
检测模块，用于检测到待发送消息的标签为非实时发送的消息标签；
确定模块，用于针对所述待发送消息的用户设备群，根据消息及时响应率以及修正策略动态确定所述用户设备群中每个用户设备响应消息的最佳时间段；
发送模块，用于根据每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述用户设备群中对应的用户设备。
可选地，所述装置还包括：
预设模块，用于预设统计消息及时响应率的时间间隔；
统计模块，用于根据所述预设模块预设的时间间隔，统计每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量；
所述确定模块，用于根据所述统计模块统计的每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量，确定每个用户设备在每个时间间隔内的消息及时响应率；
所述确定模块，用于将所述消息及时响应率最大值对应的时间间隔确定为所述最佳时间段；
修正模块，用于利用预设的修正策略对所述确定模块确定的每个用户设备在每个时间间隔内的消息及时响应率进行修正，以使所述消息及时响应率的波动幅度最小化；
所述确定模块，用于将所述修正模块修正后的消息及时响应率的最大值对应的时间间隔确定为所述最佳时间段。
可选地，若所述用户设备的消息及时响应率最大值包括多个时，所述确定模块还用于：
分别获取多个所述消息及时响应率最大值对应的时间间隔，将时间最早的时间间隔确定为所述用户设备响应消息的最佳时间段；或者将网络状态最优的时间间隔确定为所述用户设备响应消息的最佳时间段。
可选地，所述的装置还包括：
分组模块，用于根据预设的比例，将所述用户设备群随机分为第一用户设备群和第二用户设备群；
所述发送模块，还用于根据所述确定模块确定的所述第一用户设备群中每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述第一用户设备群中对应的用户设备；
所述发送模块，还用于将所述待发送消息随机发送给所述第二用户设备群中的每个用户设备。
可选地，所述统计模块，还用于统计所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率；
所述装置还包括：
比对模块，用于当比对得到所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率之间的差距小于等于预设的差距阈值，则重新通过所述确定模块确定用户设备群中每个用户设备在所述最佳时间段。
可选地，所述确定模块，还用于当确定所述用户设备群中存在用户设备在所述最佳时间段的消息及时响应率低于预设的消息及时响应率阈值时，重新确定所述用户设备响应消息的最佳时间段。
可选地，所述检测模块，还用于检测到待发送消息的标签为实时发送的 消息标签；
所述发送模块，还用于将所述待发送消息实时发送给所述用户设备群中每个用户设备。
本申请还提供一种应用后台服务器，包括：上述第二方面所述的应用消息推送装置。
本发明实施例通过根据待发送消息的用户设备群，确定所述用户设备群中每个用户设备响应消息的最佳时间段，即可以获取到每个用户点击消息的最大概率时间段，根据每个用户设备响应消息的最佳时间段(即用户点击消息的最大概率时间段)，将所述待发送消息分别发送给所述用户设备群中对应的用户设备，使得不同用户在不同时间点收到***的推送消息，从而最大化用户点击消息的概率。由于本发明充分考虑用户点击消息的时间点偏好，个性化定制用户接收消息的时间点，通过个性化的时间点推送消息，从而最大化用户点击消息的概率，大大提高了消息命中目标用户的有效性。因此可以解决现有技术中存在的批量消息推送方法中存在消息及时响应率较低的问题。
本发明还可以考虑到用户点击消息的时间点偏好发生变化，并进行自适应地修正用户设备响应消息的最佳时间段，保证了消息命中目标用户的有效性。
当然，实施本申请的任一产品必不一定需要同时达到以上所述的所有技术效果。
附图说明
此处所说明的附图用来提供对本申请的进一步理解，构成本申请的一部分，本申请的示意性实施例及其说明用于解释本申请，并不构成对本申请的不当限定。在附图中：
图1是本申请实施例的一种应用消息推送方法的流程示意图；
图2是图1所示实施例一中步骤102的一种实现方法流程示意图；
图3是图1所示实施例一中步骤102的又一种实现方法流程示意图；
图4是本申请实施例的一种应用消息推送方法的流程示意图；
图5是本申请实施例的一种应用消息推送方法的具体实现功能示意图；
图6为本申请实施例提供的一种应用消息推送装置的结构示意图；
图7为本申请实施例提供的一种应用后台服务器的结构示意图。
具体实施方式
以下将配合附图及实施例来详细说明本申请的实施方式，藉此对本申请如何应用技术手段来解决技术问题并达成技术功效的实现过程能充分理解并据以实施。
为了解决消息及时响应率较低的问题，在现有技术中，将一天划分为若干个时间段，根据每一个时间段内用户浏览消息的次数，对不同的时间段内的浏览次数做比较；将浏览次数最多的时间段设置为优先权最高的时间段，进一步地，判断终端接收消息的时间是否落在优先权最高时间段内，若是，则实时推送接收消息给用户，否则推迟推送消息给用户。然而发明人在实现本发明的过程中发现现有技术中存在如下问题：
首先，在设置优先权最高的时间段时，只是将用户最多一次浏览次数的时间段作为最优时间段，但是用户浏览的前提取决于现在有没有收到过消息，很有可能用户现在很闲，但是终端没有消息，用户也就不会去浏览，所以现有技术中，优先权最高的时间段的设置并不能准确反映用户点击消息的偏好。而且现有技术中，优先权最高的时间段一旦设置，用户只会在这个优先权最高的时间段收到终端推送的接收消息，当用户习惯发生改变时，不能进行自适应的调整优先权最高的时间段。
其次，现有技术中，是终端(如手机)接收各个应用后台或操作***产生的消息，之后根据上述设置的优先权最高的时间段，将接收消息推送给用户。因此，推送消息是终端做的，但是，由于所有消息产生都不是操作***本身，还包括各个应用后台产生的消息，终端是很难区分实时性要求很高和实时性要求不高的消息，比如聊天消息，突发新闻、消费的动账提醒等实时性要求很高的笑死是需要终端立即推送给用户的，在现有技术中，优先权最高的时间段一旦设置，用户只会在这个优先权最高的时间段收到终端推送的 接收消息，这样实时性要求很高的消息不能实时推送给用户，会造成用户漏掉重要的实时性消息。
因此，现有技术中存在优先权最高的时间段的设置缺少自适应调整机制(比较机械刻板)，并不能准确反映用户点击消息的偏好，从而导致消息命中目标用户的有效性不能得到有效提高，其次，会造成用户漏掉重要的实时性消息。
本发明为了解决上述现有技术中存在的问题，通过自适应调整机制，个性化调整每个用户点击消息的时间点偏好，从而实现个性化定制用户接收消息的时间点，通过个性化定制的用户接收消息的时间点向用户推送消息，从而最大化用户点击消息的概率，提高消息命中目标用户的有效性。
实施例一
图1是本申请实施例的一种应用消息推送方法的流程示意图；如图1所示，包括：
101、检测待发送消息是否为非实时发送的消息；
本发明实施例中，每个消息携带有消息标签，其中，消息标签包括非实时发送的消息标签和实时发送的消息标签。为此，根据消息中携带的消息标签，当消息标签为非实时发送的消息标签，可以确定待发送消息为非实时发送的消息，当消息标签为实时发送的消息标签，可以确定待发送消息为实时发送的消息。
当检测到待发送消息是非实时发送的消息时，执行步骤102，当检测到待发送消息是实时发送的消息时，执行步骤104。
102、针对所述待发送消息的用户设备群，根据消息及时响应率以及修正策略动态确定所述用户设备群中每个用户设备响应消息的最佳时间段；
本发明实施例中的用户设备例如可以为智能手机、PAD(平板电脑)、PC平板二合一等，其中，所述用户设备可以具有有线网络接口，也可以具有无线网络接口。
举例来说，例如手机淘宝(手机中的一个应用APP)的应用后台需要批量发送商品促销消息给用户手机，由于安装有手机淘宝的用户手机数量很大，这里的用户手机属于一种用户设备，因此，本发明实施例中称为用户设 备群。
为了能够实现最大化用户点击商品促销消息的概率，本发明实施例需要获取用户群中每个用户点击消息的时间点偏好。需要说明的是，当用户设备接收到手机淘宝的应用后台发送的商品促销消息时，用户通过用户设备点击商品促销消息也可以称为用户设备响应该商品促销消息，因此，步骤101中确定所述用户设备群中每个用户设备响应消息的最佳时间段其实质就是获取用户群中每个用户点击消息的时间点偏好。
目前的手机批量消息推送(不包括事件规则实时触发的消息)，都是开始于某个指定的时间点T，真正触达到用户只取决于时间点T和***推送处理的时间。由于用户点击消息的概率随着收到消息的时间呈指数衰减，即绝大多数用户只会在收到消息很短的时间内(一般是半小时内)点击，因此用户是否点击消息与用户收到的时间点是强相关的。本发明考虑用户收到消息的时间点偏好，可以对批量推送消息让不同用户在不同时间点收到，即个性化调整每个用户点击消息的时间点偏好，从而实现个性化定制用户接收消息的时间点，通过个性化定制的用户接收消息的时间点向用户推送消息，从而最大化用户点击消息的概率，提高消息命中目标用户的有效性。
具体地，本发明实施例中例如可以通过统计每一个手机淘宝用户(相当于安装有手机淘宝的用户设备)历史(例如最近1年)收到消息和点击消息的情况，从而分析出每一个手机淘宝用户点击消息的最大概率时间段(相当于响应消息的最佳时间段)。
步骤102在具体实现时可以参考图2和图3所示实施例中的详细描述。
103、根据每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述用户设备群中对应的用户设备。
基于上述分析得到的每一个手机淘宝用户点击消息的最大概率时间段(相当于响应消息的最佳时间段)，手机淘宝的应用后台将商品促销消息在对应的最大概率时间段(最佳时间段)内分别发送给对应的手机淘宝用户。举例来说，假设分析得到用户A点击手机淘宝的应用后台发送的商品促销消息的最大概率时间段(最佳时间段)是中午12点到12点10分之间，则手机淘宝的应用后台在中午12点到12点10分之间发送商品促销消息给用户A使用的用户设备，这样，用户A在接收到消息的时候会很快点击该消息，这 样，消息命中目标用户的有效性就得到提高。
104、将所述待发送消息实时发送给所述用户设备群中每个用户设备。
本发明实施例通过根据待发送消息的用户设备群，确定所述用户设备群中每个用户设备响应消息的最佳时间段，即可以获取到每个用户点击消息的最大概率时间段，根据每个用户设备响应消息的最佳时间段(即用户点击消息的最大概率时间段)，将所述待发送消息分别发送给所述用户设备群中对应的用户设备，使得不同用户在不同时间点收到***的推送消息，从而最大化用户点击消息的概率。由于本发明充分考虑用户点击消息的时间点偏好，个性化定制用户接收消息的时间点，通过个性化的时间点推送消息，从而最大化用户点击消息的概率，大大提高了消息命中目标用户的有效性。因此可以解决现有技术中存在的批量消息推送方法中存在消息及时响应率较低的问题。
图2是图1所示实施例一中步骤102的一种实现方法流程示意图，如图2所示，本发明实施例中，步骤102中确定所述用户设备群中每个用户设备响应消息的最佳时间段具体可以包括如下：
201、预设统计消息及时响应率的时间间隔；
为了可以更准确地确定响应消息的最佳时间段，本发明实施例中可以将统计消息及时响应率的时间间隔设置的小一点，例如，将每天按照固定的时间间隔Δt分成n个时间间隔，比如以15分钟为时间间隔，就将1天切分成96个时间区分，此时Δt＝15分钟，n＝96个。
202、根据预设的时间间隔，统计每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量；
本步骤主要是统计用户历史收到消息和点击消息的情况，为了更加准确地确定响应消息的最佳时间段，统计的样本(即用户历史收到消息和点击消息的情况)多一点比较好，例如可以统计用户最近1年收到消息和点击消息的情况，具体统计的时候，是统计每个用户设备在最近1年中，在每个时间间隔内接收到的消息数量和及时响应的消息数量。
由于用户在工作日和节假日接收消息的数量和点击消息的数量以及点击消息的时间段也是不同的，所以在统计的时候可以区分工作日和节假日从而进行对应的区别统计。
203、根据每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量，确定每个用户设备在每个时间间隔内的消息及时响应率；
根据步骤202中统计得到的每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量，将每个用户设备在每个时间间隔内的及时响应的消息数量除以接收到的消息数量就可以得到该用户设备在该时间间隔内的消息及时响应率，因此可以得到每个用户设备在每个时间间隔内的消息及时响应率。
204、将所述消息及时响应率最大值对应的时间间隔确定为所述最佳时间段。
根据步骤203中得到的每个用户设备在每个时间间隔内的消息及时响应率，通常，某一个时间间隔内的消息及时响应率高说明该时间间隔内用户在接收到消息时及时点击该消息的概率就高，因此，可以将消息及时响应率最大值对应的时间间隔确定为用户设备响应消息的最佳时间段。
图3是图1所示实施例一中步骤102的又一种实现方法流程示意图，如图3所示，本发明实施例中，考虑到用户设备在各个时间间隔内收到的消息数量可能差异较大，造成每个时间间隔内的消息及时响应率的波动幅度较大，从而影响准确判断每个用户设备响应消息的最佳时间段，为此，本发明实施例中需要对每个时间间隔内的消息及时响应率进行修正；具体地，步骤102可以包括如下：
301、预设统计消息及时响应率的时间间隔；
为了可以更准确地确定响应消息的最佳时间段，本发明实施例中可以将统计消息及时响应率的时间间隔设置的小一点，例如，将每天按照固定的时间间隔Δt分成n个时间间隔，比如以15分钟为时间间隔，就将1天切分成96个时间区分，此时Δt＝15分钟，n＝96个。
302、根据预设的时间间隔，统计每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量；
本步骤主要是统计用户历史收到消息和点击消息的情况，为了更加准确地确定响应消息的最佳时间段，统计的样本(即用户历史收到消息和点击消息的情况)多一点比较好，例如可以统计用户最近1年收到消息和点击消息的情况，具体统计的时候，是统计每个用户设备在最近1年中，在每个时间 间隔内接收到的消息数量和及时响应的消息数量。
由于用户在工作日和节假日接收消息的数量和点击消息的数量以及点击消息的时间段也是不同的，所以在统计的时候可以区分工作日和节假日从而进行对应的区别统计。
303、根据每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量，确定每个用户设备在每个时间间隔内的消息及时响应率；
根据步骤302中统计得到的每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量，将每个用户设备在每个时间间隔内的及时响应的消息数量除以接收到的消息数量就可以得到该用户设备在该时间间隔内的消息及时响应率，因此可以得到每个用户设备在每个时间间隔内的消息及时响应率。
304、利用预设的修正策略对所述确定的每个用户设备在每个时间间隔内的消息及时响应率进行修正，以使每个时间间隔内的消息及时响应率的波动幅度最小化；
本发明实施例中，预设的修正策略例如可以采用威尔逊区间进行修正，例如，用户在时间间隔1收到消息10条，点击6条，在时间间隔2收到消息100条，点击60条，计算的点击概率都是为0.6，经过威尔逊区间修正后，用户在时间间隔1和时间间隔2的点击概率分别为0.31和0.50(置信度为0.95的正态分布置信区间的下限)，即认为时间间隔2通过频率的方式计算的点击概率更稳定，波动幅度更小。
305、将所述修正后的消息及时响应率的最大值对应的时间间隔确定为所述最佳时间段。
需要说明的是，假设上述图2和图3中，用户设备的消息及时响应率最大值包括多个时，为了可以更准确地确定用户设备响应消息的最佳时间段，本发明实施例中，例如分别获取多个所述消息及时响应率最大值对应的时间间隔，将时间最早的时间间隔确定为用户设备响应消息的最佳时间段；又例如还可以将网络状态最优的时间间隔确定为用户设备响应消息的最佳时间段。
实施例二
图4是本申请实施例的一种应用消息推送方法的流程示意图，如图4所 示，考虑到用户点击消息的时间点偏好发生变化，为了保证消息命中目标用户的有效性，本发明实施例采用对用户设备群按照预设的比例进行随机分组，针对其中一组用户设备群，采用在响应消息的最佳时间段推送消息的方式，针对另一组用户设备群，采用随机选择时间点推送消息的方式，然后对两组的用户设备响应消息的效果进行比对，根据效果比对或者两组用户设备对消息推送时间的反馈信息，判断是否对用户设备响应消息的最佳时间段进行重新计算，具体实现包括如下：
401、根据预设的比例，将所述用户设备群随机分为第一用户设备群和第二用户设备群；
具体到某个用户A在每次推送时既可能被分到第一用户设备群,也可能被分到第二用户设备群；
402、确定第一用户设备群中每个用户设备响应消息的最佳时间段；
403、根据所述第一用户设备群中每个用户设备响应消息的最佳时间段，将待发送消息分别发送给所述第一用户设备群中对应的用户设备，将待发送消息随机发送给所述第二用户设备群中的每个用户设备；
404、统计和比对所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率；
405、当比对得到所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率之间的差距小于等于预设的差距阈值，则重新确定用户设备群中每个用户设备在所述最佳时间段。
为此，本发明实施例中还可以预设消息及时响应率的差距阈值，当比对得到所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率之间的差距小于等于预设的差距阈值时，说明上述统计得到的用户设备响应消息的最佳时间段存在问题，需要重新确定。
需要说明的是，本发明实施例中还可以预设消息及时响应率阈值，当确定所述用户设备群中存在用户设备在所述最佳时间段的消息及时响应率低于预设的消息及时响应率阈值时，需要重新确定该用户设备响应消息的最佳 时间段，说明该用户设备的用户点击消息的时间点偏好可能发生了变化，为了确保在用户点击消息的时间点偏好发生变化时，本发明实施例可以及时调整用户设备响应消息的最佳时间段，保证用户设备响应消息的最佳时间段的准确性。
本发明实施例通过对用户设备群按照预设的比例进行随机分组，针对其中一组用户设备群，采用在响应消息的最佳时间段推送消息的方式，针对另一组用户设备群，采用随机选择时间点推送消息的方式，然后对两组的用户设备响应消息的效果进行比对，根据效果比对或者两组用户设备对消息推送时间的反馈信息，判断是否对用户设备响应消息的最佳时间段进行重新计算，可以考虑到用户点击消息的时间点偏好发生变化，并进行自适应地修正用户设备响应消息的最佳时间段，保证了消息命中目标用户的有效性。
实施例三
图5是本申请实施例的一种应用消息推送方法的具体实现功能示意图，如图5所示，包括：
步骤1：用户时间点偏好计算引擎；
通过用户历史(最近1年)收到消息和点击消息情况，计算用户点击消息的最大概率时间间隔，具体方法如下：
(1)区分工作日和节假日，将每天按照固定的时间间隔Δt分成n个时间间隔，比如以15分钟为时间间隔，就将1天切分成96个时间区分，此时Δt＝15，n＝96；
(2)区分工作日和节假日，计算用户历史上(这里取最近1年)在每个时间间隔上收到消息数和点击数，计算点击概率；
(3)考虑到用户在各个时间间隔上收到的消息数可能差异较大，利用威尔逊区间进行修正，例如用户在时间间隔1收到消息10条，点击6条，在时间间隔2收到消息100条，点击60条，计算的点击概率同为0.6，经过威尔逊区间修正后，用户在时间间隔1和时间间隔2的点击概率分别为0.31和0.50(置信度为0.95的正态分布置信区间的下限)，即认为时间间隔2通过频率的方式计算的点击概率更稳定，波动幅度更小。
(4)区分工作日和节假日，选取每个用户修正后的点击概率最大的时间间隔，剔除只有m(m≥1)个消息区间发送数的用户以及在所有消息区间点 击数都为0的用户(无法判断用户的偏好时间点)；对于概率最大的时间间隔存在多个的，选取最早的那个时间间隔。
(5)考虑从***发送到触达用户的平均延迟Δt′(如果很小可忽略)，将计算好的用户偏好时间点减去Δt′作为该用户的推送时间点，并以数据表的形式推送到消息发送装置存储，其中，用户偏好时间点例如可以在确定的最佳时间段中取一个时间点。
步骤2：选择发送内容及用户群
发送内容可以由业务人员指定或者***计算得到，用户群可以是全量用户也可以是通过标签筛选的用户，比如18-22岁用户，女性用户等等。
步骤3：消息发送
发送的内容和用户群推送到消息发送装置，选择用户最佳的推送时间点后，在用户最佳的推送时间点发送给用户，具体包括：
(1)用户群分组装置：后台异步线程将步骤2中指定的用户群，结合步骤1中计算出的用户时间点偏好信息，将用户以步骤1中提到的时间间隔Δt进行分组。对于没有明确时间点偏好信息的用户,将其随机分配至一个发送时间分组内。
(2)分桶装置：用户分组结束后，经过分桶装置，将一定比例的用户(比如20％)发送时间置为随机发送时间，用于效果对比同时收集更为丰富的用户对消息发送时间的反馈信息。
(3)用户消息发送装置：用户消息发送装置，会通过分布式的并发调度控制，在多台集群机器上按照Δt时间间隔定时启动任务执行线程，并发查询(2)中分桶装置在该***时间段内用户发送数据，并发触发消息推送。这里，为避免分布式***造成的同一用户的发送任务重叠(例如:A,B,C三台机器都在同一时间点拿到了针对用户1的发送任务，如果不做处理，A,B,C会分别触发一次针对用户1的消息推送)，***会直接做根据用户ID维度和执行任务机器的逻辑分组，保证一个用户只有一台机器负责消息推送。为此，本发明实施例中需要根据所述用户设备群中每个用户设备的标识，确定与每个用户设备的标识对应的发送服务器的标识，以使所述发送服务器将所述待发送消息发送给对应的用户设备；具体做法如下：
a.将任务执行机器进行逻辑编号，起始编号为0。例如：有20台机器执行任务，那么逻辑编号是0-19的所有整数。
b.将用户的ID信息针对机器编号进行映射处理。以上面的例子为例，假设用户1的ID为123，那么对123针对20取模得到3，意味着用户1需要由逻辑编号为3的机器负责消息推送。
c.所有机器并发执行该***时间段内的待推送用户数据获取，但是只处理根据b中提到的算法计算之后属于自己职责范围内的用户数据,在提高推送并发度的同时避免任务重叠。
本发明实施例充分考虑用户点击消息的时间点偏好，个性化定制用户接收消息的时间点，通过个性化的时间点推送消息，从而最大化用户点击消息的概率。同时，也考虑到用户点击消息的时间点偏好发生变化，并进行自适应地修正用户设备响应消息的最佳时间段，保证了消息命中目标用户的有效性。
图6为本申请实施例提供的一种应用消息推送装置的结构示意图，如图6所示，包括：
检测模块60，用于检测到待发送消息的标签为非实时发送的消息标签；
确定模块61，用于针对所述待发送消息的用户设备群，根据消息及时响应率以及修正策略动态确定所述用户设备群中每个用户设备响应消息的最佳时间段；
发送模块62，用于根据每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述用户设备群中对应的用户设备。
其中，所述装置还包括：
预设模块63，用于预设统计消息及时响应率的时间间隔；
统计模块64，用于根据所述预设模块预设的时间间隔，统计每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量；
所述确定模块61，用于根据所述统计模块统计的每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量，确定每个用户设备在每个时间间隔内的消息及时响应率；
所述确定模块61，用于将所述消息及时响应率最大值对应的时间间隔确定为所述最佳时间段。
其中，所述的装置还包括：
修正模块65，用于利用预设的修正策略对所述确定模块确定的每个用户设备在每个时间间隔内的消息及时响应率进行修正，以使所述消息及时响应率的波动幅度最小化；
所述确定模块61，用于将所述修正模块修正后的消息及时响应率的最大值对应的时间间隔确定为所述最佳时间段。
若所述用户设备的消息及时响应率最大值包括多个时，所述确定模块61还用于：
分别获取多个所述消息及时响应率最大值对应的时间间隔，将时间最早的时间间隔确定为所述用户设备响应消息的最佳时间段；或者将网络状态最优的时间间隔确定为所述用户设备响应消息的最佳时间段。
其中，所述的装置还包括：
分组模块66，用于根据预设的比例，将所述用户设备群随机分为第一用户设备群和第二用户设备群；
所述发送模块62，还用于根据所述确定模块确定的所述第一用户设备群中每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述第一用户设备群中对应的用户设备；
所述发送模块62，还用于将所述待发送消息随机发送给所述第二用户设备群中的每个用户设备。
其中，所述统计模块64，还用于统计所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率。
所述装置还包括：
比对模块67，用于当比对得到所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率之间的差距小于等于预设的差距阈值，则重新通过所述确定模块61确定用户设备群中每个用户设备在所述最佳时间段。
其中，所述确定模块61，还用于当确定所述用户设备群中存在用户设备在所述最佳时间段的消息及时响应率低于预设的消息及时响应率阈值时，重新确定所述用户设备响应消息的最佳时间段。
所述检测模块60，还用于检测到待发送消息的标签为实时发送的消息标签；
所述发送模块62，还用于将所述待发送消息实时发送给所述用户设备群中每个用户设备
本发明实施例的装置充分考虑用户点击消息的时间点偏好，个性化定制用户接收消息的时间点，通过个性化的时间点推送消息，从而最大化用户点击消息的概率。同时，也考虑到用户点击消息的时间点偏好发生变化，并进行自适应地修正用户设备响应消息的最佳时间段，保证了消息命中目标用户的有效性。
图7为本申请实施例提供的一种应用后台服务器的结构示意图，其中，所述后台服务器可以具有有线网络接口，也可以具有无线网络接口。如图7所示，包括：存储器10；一个或多个处理器11；
其中，存储器10中保存有实现上述图1-图5中任一实施例中所述的消息推送方法的指令，当处理器11调用存储器10中保存的指令，可以执行以下步骤：
检测到待发送消息的标签为非实时发送的消息标签；
针对所述待发送消息的用户设备群，根据消息及时响应率以及修正策略动态确定所述用户设备群中每个用户设备响应消息的最佳时间段；
根据每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述用户设备群中对应的用户设备。
其中，确定所述用户设备群中每个用户设备响应消息的最佳时间段包括：
预设统计消息及时响应率的时间间隔；
根据预设的时间间隔，统计每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量；
根据每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量，确定每个用户设备在每个时间间隔内的消息及时响应率；
将所述消息及时响应率最大值对应的时间间隔确定为所述最佳时间段。
其中，确定每个用户设备在每个时间间隔内的消息及时响应率之后，包括：
利用预设的修正策略对所述确定的每个用户设备在每个时间间隔内的消息及时响应率进行修正，以使所述消息及时响应率的波动幅度最小化；
对应地，将所述修正后的消息及时响应率的最大值对应的时间间隔确定为所述最佳时间段。
可选地，若所述用户设备的消息及时响应率最大值包括多个时，处理器11还可以执行如下步骤：
分别获取多个所述消息及时响应率最大值对应的时间间隔，将时间最早的时间间隔确定为所述用户设备响应消息的最佳时间段；或者将网络状态最优的时间间隔确定为所述用户设备响应消息的最佳时间段。
可选地，处理器11还可以执行如下步骤：
根据预设的比例，将所述用户设备群分为第一用户设备群和第二用户设备群；
根据所述第一用户设备群中每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述第一用户设备群中对应的用户设备；
将所述待发送消息随机发送给所述第二用户设备群中的每个用户设备；
统计和比对所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率。
可选地，处理器11还可以执行如下步骤：
当确定所述用户设备群中存在用户设备在所述最佳时间段的消息及时响应率低于预设的消息及时响应率阈值时，重新确定所述用户设备响应消息的最佳时间段。
其中，将所述待发送消息分别发送给所述用户设备群中对应的用户设备包括：
根据所述用户设备群中每个用户设备的标识，确定与每个用户设备的标识对应的发送服务器的标识，以使所述发送服务器将所述待发送消息发送给对应的用户设备。
可选地，处理器11还可以执行如下步骤：
检测到待发送消息的标签为实时发送的消息标签；
将所述待发送消息实时发送给所述用户设备群中每个用户设备。
本发明实施例的后台服务设备充分考虑用户点击消息的时间点偏好，个性化定制用户接收消息的时间点，通过个性化的时间点推送消息，从而最大化用户点击消息的概率。同时，也考虑到用户点击消息的时间点偏好发生变化，并进行自适应地修正用户设备响应消息的最佳时间段，保证了消息命中目标用户的有效性。
在一个典型的配置中，计算设备包括一个或多个处理器(CPU)、输入/输出接口、网络接口和内存。
内存可能包括计算机可读介质中的非永久性存储器，随机存取存储器(RAM)或非易失性内存等形式，如只读存储器(ROM)或闪存(flash RAM)。内存是计算机可读介质的示例。
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括，但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带，磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质，可用于存储可以被计算设备访问的信息。按照本文中的界定，计算机可读介质不包括非暂存电脑可读媒体(transitory media)，如调制的数据信号和载波。
如在说明书及权利要求当中使用了某些词汇来指称特定组件。本领域技术人员应可理解，硬件制造商可能会用不同名词来称呼同一个组件。本说明书及权利要求并不以名称的差异来作为区分组件的方式，而是以组件在功能上的差异来作为区分的准则。如在通篇说明书及权利要求当中所提及的“包含”为一开放式用语，故应解释成“包含但不限定于”。“大致”是指在可接收的误差范围内，本领域技术人员能够在一定误差范围内解决所述技术问题，基本达到所述技术效果。此外，“耦接”一词在此包含任何直接及间接的电性耦接手段。因此，若文中描述一第一装置耦接于一第二装置，则代表所述第一装置可直接电性耦接于所述第二装置，或通过其他装置或耦接手段 间接地电性耦接至所述第二装置。说明书后续描述为实施本申请的较佳实施方式，然所述描述乃以说明本申请的一般原则为目的，并非用以限定本申请的范围。本申请的保护范围当视所附权利要求所界定者为准。
还需要说明的是，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的商品或者***不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种商品或者***所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的商品或者***中还存在另外的相同要素。
上述说明示出并描述了本发明的若干优选实施例，但如前所述，应当理解本发明并非局限于本文所披露的形式，不应看作是对其他实施例的排除，而可用于各种其他组合、修改和环境，并能够在本文所述发明构想范围内，通过上述教导或相关领域的技术或知识进行改动。而本领域人员所进行的改动和变化不脱离本发明的精神和范围，则都应在本发明所附权利要求的保护范围内。

Claims (16)

  1. 一种应用消息推送方法，位于应用后台服务器侧执行，其特征在于，包括：
    检测到待发送消息的标签为非实时发送的消息标签；
    针对所述待发送消息的用户设备群，根据消息及时响应率以及修正策略动态确定所述用户设备群中每个用户设备响应消息的最佳时间段；
    根据每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述用户设备群中对应的用户设备。
  2. 根据权利要求1所述的方法，其特征在于，根据消息及时响应率以及修正策略动态确定所述用户设备群中每个用户设备响应消息的最佳时间段包括：
    预设统计消息及时响应率的时间间隔；
    根据预设的时间间隔，统计每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量；
    根据每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量，确定每个用户设备在每个时间间隔内的消息及时响应率；
    将所述消息及时响应率最大值对应的时间间隔确定为所述最佳时间段；
    利用预设的修正策略对所述确定的每个用户设备在每个时间间隔内的消息及时响应率进行修正，以使所述消息及时响应率的波动幅度最小化；
    对应地，将所述修正后的消息及时响应率的最大值对应的时间间隔确定为所述最佳时间段。
  3. 根据权利要求2所述的方法，其特征在于，若所述用户设备的消息及时响应率最大值包括多个时，所述方法还包括：
    分别获取多个所述消息及时响应率最大值对应的时间间隔，将时间最早的时间间隔确定为所述用户设备响应消息的最佳时间段；或者将网络状态最优的时间间隔确定为所述用户设备响应消息的最佳时间段。
  4. 根据权利要求1所述的方法，其特征在于，还包括：
    根据预设的比例，将所述用户设备群分为第一用户设备群和第二用户设备群；
    根据所述第一用户设备群中每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述第一用户设备群中对应的用户设备；
    将所述待发送消息随机发送给所述第二用户设备群中的每个用户设备。
  5. 根据权利要求4所述的方法，其特征在于，还包括：
    统计所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率；
    当比对得到所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率之间的差距小于等于预设的差距阈值，则重新确定所述用户设备群中每个用户设备在所述最佳时间段。
  6. 根据权利要求1所述的方法，其特征在于，还包括：
    当确定所述用户设备群中存在用户设备在所述最佳时间段的消息及时响应率低于预设的消息及时响应率阈值时，重新确定所述用户设备响应消息的最佳时间段。
  7. 根据权利要求1所述的方法，其特征在于，还包括：
    检测到待发送消息的标签为实时发送的消息标签；
    将所述待发送消息实时发送给所述用户设备群中每个用户设备。
  8. 根据权利要求1所述的方法，其特征在于，将所述待发送消息分别发送给所述用户设备群中对应的用户设备包括：
    根据所述用户设备群中每个用户设备的标识，确定与每个用户设备的标识对应的发送服务器的标识，以使所述发送服务器将所述待发送消息发送给对应的用户设备。
  9. 一种应用消息推送装置，位于应用后台服务器侧，其特征在于，包括：
    检测模块，用于检测到待发送消息的标签为非实时发送的消息标签；
    确定模块，用于针对所述待发送消息的用户设备群，根据消息及时响应率以及修正策略动态确定所述用户设备群中每个用户设备响应消息的最佳时间段；
    发送模块，用于根据每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述用户设备群中对应的用户设备。
  10. 根据权利要求9所述的装置，其特征在于，所述装置还包括：
    预设模块，用于预设统计消息及时响应率的时间间隔；
    统计模块，用于根据所述预设模块预设的时间间隔，统计每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量；
    所述确定模块，用于根据所述统计模块统计的每个用户设备在每个时间间隔内接收到的消息数量和及时响应的消息数量，确定每个用户设备在每个时间间隔内的消息及时响应率；
    所述确定模块，用于将所述消息及时响应率最大值对应的时间间隔确定为所述最佳时间段；
    修正模块，用于利用预设的修正策略对所述确定模块确定的每个用户设备在每个时间间隔内的消息及时响应率进行修正，以使所述消息及时响应率的波动幅度最小化；
    所述确定模块，用于将所述修正模块修正后的消息及时响应率的最大值对应的时间间隔确定为所述最佳时间段。
  11. 根据权利要求10所述的装置，其特征在于，若所述用户设备的消息及时响应率最大值包括多个时，所述确定模块还用于：
    分别获取多个所述消息及时响应率最大值对应的时间间隔，将时间最早的时间间隔确定为所述用户设备响应消息的最佳时间段；或者将网络状态最优的时间间隔确定为所述用户设备响应消息的最佳时间段。
  12. 根据权利要求9所述的装置，其特征在于，还包括：
    分组模块，用于根据预设的比例，将所述用户设备群随机分为第一用户设备群和第二用户设备群；
    所述发送模块，还用于根据所述确定模块确定的所述第一用户设备群中每个用户设备响应消息的最佳时间段，将所述待发送消息分别发送给所述第一用户设备群中对应的用户设备；
    所述发送模块，还用于将所述待发送消息随机发送给所述第二用户设备群中的每个用户设备。
  13. 根据权利要求12所述的装置，其特征在于：
    所述统计模块，还用于统计所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的 消息及时响应率；
    所述装置还包括：
    比对模块，用于当比对得到所述第一用户设备群中每个用户设备在所述最佳时间段的消息及时响应率和所述第二用户设备群中的每个用户设备的消息及时响应率之间的差距小于等于预设的差距阈值，则重新通过所述确定模块确定用户设备群中每个用户设备在所述最佳时间段。
  14. 根据权利要求9所述的装置，其特征在于：
    所述确定模块，还用于当确定所述用户设备群中存在用户设备在所述最佳时间段的消息及时响应率低于预设的消息及时响应率阈值时，重新确定所述用户设备响应消息的最佳时间段。
  15. 根据权利要求9所述的装置，其特征在于：
    所述检测模块，还用于检测到待发送消息的标签为实时发送的消息标签；
    所述发送模块，还用于将所述待发送消息实时发送给所述用户设备群中每个用户设备。
  16. 一种应用后台服务器，其特征在于包括：
    如权利要求9-15中任一项所述的应用消息推送装置。
PCT/CN2017/093534 2016-07-22 2017-07-19 一种应用消息推送方法、装置 WO2018014846A1 (zh)

Priority Applications (9)

Application Number Priority Date Filing Date Title
SG11201900503VA SG11201900503VA (en) 2016-07-22 2017-07-19 Method and device for pushing application message
EP17830480.4A EP3490198B1 (en) 2016-07-22 2017-07-19 Method and device for pushing application message
PL17830480T PL3490198T3 (pl) 2016-07-22 2017-07-19 Sposób i urządzenie do wymuszonego przekazywania wiadomości aplikacji
JP2019503319A JP6765500B2 (ja) 2016-07-22 2017-07-19 アプリケーションメッセージをプッシュする方法及びデバイス
KR1020197005197A KR102192877B1 (ko) 2016-07-22 2017-07-19 애플리케이션 메시지를 푸싱하기 위한 방법 및 디바이스
ES17830480T ES2818610T3 (es) 2016-07-22 2017-07-19 Método y dispositivo para enviar un mensaje de aplicación
EP20189068.8A EP3751797B1 (en) 2016-07-22 2017-07-19 Method and device for pushing application message
US16/254,329 US10812607B2 (en) 2016-07-22 2019-01-22 Method and device for pushing application message
US16/723,676 US10958749B2 (en) 2016-07-22 2019-12-20 Method and device for pushing application message

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610584814.3 2016-07-22
CN201610584814.3A CN106899488B (zh) 2016-07-22 2016-07-22 一种应用消息推送方法、装置

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/254,329 Continuation US10812607B2 (en) 2016-07-22 2019-01-22 Method and device for pushing application message

Publications (1)

Publication Number Publication Date
WO2018014846A1 true WO2018014846A1 (zh) 2018-01-25

Family

ID=59190971

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/093534 WO2018014846A1 (zh) 2016-07-22 2017-07-19 一种应用消息推送方法、装置

Country Status (10)

Country Link
US (2) US10812607B2 (zh)
EP (2) EP3490198B1 (zh)
JP (1) JP6765500B2 (zh)
KR (1) KR102192877B1 (zh)
CN (1) CN106899488B (zh)
ES (1) ES2818610T3 (zh)
PL (1) PL3490198T3 (zh)
SG (1) SG11201900503VA (zh)
TW (1) TWI736633B (zh)
WO (1) WO2018014846A1 (zh)

Families Citing this family (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106899488B (zh) 2016-07-22 2020-09-15 阿里巴巴集团控股有限公司 一种应用消息推送方法、装置
US10505871B1 (en) * 2017-11-30 2019-12-10 Sandeep Jain Future messaging maximizing contextual relevancy and minimizing information overload based distractions
CN109995798B (zh) * 2017-12-29 2021-05-04 Tcl科技集团股份有限公司 一种信息推送的方法及其设备
CN108400929B (zh) * 2018-02-08 2021-06-08 网易乐得科技有限公司 数据处理方法、装置、计算设备和介质
US11240052B2 (en) * 2018-06-19 2022-02-01 International Business Machines Corporation Facilitating communication in a collaborative environment
CN110971544B (zh) * 2018-09-30 2023-07-07 北京国双科技有限公司 一种数据发送方法及装置
CN109787884B (zh) * 2019-01-02 2021-03-12 中国联合网络通信集团有限公司 一种消息推送方法和装置
CN112217648B (zh) * 2019-07-12 2022-07-08 腾讯科技（深圳）有限公司 一种消息群发的方法以及相关装置
CN110445879A (zh) * 2019-09-11 2019-11-12 上海二三四五网络科技有限公司 一种后台统一推送信息的控制方法及装置
JP7438715B2 (ja) * 2019-10-30 2024-02-27 キヤノン株式会社 情報処理装置、制御方法及びプログラム
US11102622B1 (en) * 2020-01-20 2021-08-24 Noble Systems Corporation Best time to send limited-content text messages to parties
CN112182366A (zh) * 2020-09-15 2021-01-05 北京每日优鲜电子商务有限公司 信息推送方法、装置、电子设备和计算机可读介质
CN112733012B (zh) * 2020-12-30 2023-06-02 杭州网易再顾科技有限公司 一种执行用户触达的方法、装置、设备和介质
KR102426621B1 (ko) * 2021-09-29 2022-07-29 김민혁 애플리케이션이 설치되어 있는 클라이언트 단말에 대한 효율적인 푸시 알림 메시지의 발송을 처리하는 애플리케이션 서비스 서버 및 그 동작 방법

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102801817A (zh) * 2012-09-07 2012-11-28 深圳市学之泉集团有限公司 基于用户上下文的推送方法及装置
CN103378978A (zh) * 2012-04-16 2013-10-30 腾讯科技（深圳）有限公司 推送消息的方法和服务器
US20150222583A1 (en) * 2013-12-09 2015-08-06 Tencent Technology (Shenzhen) Company Limited Systems and Methods for Message Pushing
CN106899488A (zh) * 2016-07-22 2017-06-27 阿里巴巴集团控股有限公司 一种应用消息推送方法、装置

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090191902A1 (en) 2008-01-25 2009-07-30 John Osborne Text Scripting
KR100983545B1 (ko) * 2008-06-25 2010-09-27 주식회사 케이티 모바일 푸시 서비스를 제공하기 위한 시스템 및 그 방법
US9313800B2 (en) * 2009-06-23 2016-04-12 Nokia Technologies Oy Method and apparatus for optimizing energy consumption for wireless connectivity
CN104703146B (zh) 2013-12-09 2019-03-08 腾讯科技（深圳）有限公司 信息推送方法、客户端及***
CN104717120B (zh) 2013-12-13 2019-03-01 阿里巴巴集团控股有限公司 确定信息发送时间的方法和装置
US20160014057A1 (en) * 2014-07-08 2016-01-14 Samsung Electronics Co., Ltd. Method and system for providing dynamically customized web push messages in a wireless network
CN105099870B (zh) 2015-05-28 2019-02-15 努比亚技术有限公司 一种消息推送方法及装置
US10291741B2 (en) * 2015-10-26 2019-05-14 Facebook, Inc. Notification click-through rate (CTR) thresholds
CN105610929A (zh) * 2015-12-24 2016-05-25 北京奇虎科技有限公司 一种个性化的数据推送方法和装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103378978A (zh) * 2012-04-16 2013-10-30 腾讯科技（深圳）有限公司 推送消息的方法和服务器
CN102801817A (zh) * 2012-09-07 2012-11-28 深圳市学之泉集团有限公司 基于用户上下文的推送方法及装置
US20150222583A1 (en) * 2013-12-09 2015-08-06 Tencent Technology (Shenzhen) Company Limited Systems and Methods for Message Pushing
CN106899488A (zh) * 2016-07-22 2017-06-27 阿里巴巴集团控股有限公司 一种应用消息推送方法、装置

Also Published As

Publication number Publication date
SG11201900503VA (en) 2019-02-27
EP3490198B1 (en) 2020-09-02
EP3490198A1 (en) 2019-05-29
KR20190028549A (ko) 2019-03-18
ES2818610T3 (es) 2021-04-13
CN106899488B (zh) 2020-09-15
US20190173968A1 (en) 2019-06-06
US10812607B2 (en) 2020-10-20
EP3490198A4 (en) 2019-07-24
TWI736633B (zh) 2021-08-21
CN106899488A (zh) 2017-06-27
JP2019523499A (ja) 2019-08-22
JP6765500B2 (ja) 2020-10-07
EP3751797B1 (en) 2021-09-15
PL3490198T3 (pl) 2021-01-11
KR102192877B1 (ko) 2020-12-23
US20200128092A1 (en) 2020-04-23
TW201804826A (zh) 2018-02-01
US10958749B2 (en) 2021-03-23
EP3751797A1 (en) 2020-12-16

Similar Documents

Publication Publication Date Title
WO2018014846A1 (zh) 一种应用消息推送方法、装置
US11153726B2 (en) Techniques for enhancing group communication on a mobile device
US20170264505A1 (en) Techniques for tracking resource usage statistics per transaction across multiple layers of protocols
US20140189533A1 (en) Dynamic notification system and method
US10706069B2 (en) Techniques for replication of a client database to remote devices
WO2018121404A1 (zh) 一种超时监控方法及装置
US20150281035A1 (en) Monitoring of Node.JS Applications
US10992633B1 (en) Methods and systems for determining an unread message count
US10970139B2 (en) Controlling the triggering of function calls from content items
CN110677684B (zh) 视频处理、视频访问方法及分布式存储、视频访问***
US10855761B1 (en) Techniques for in-place directive execution
CN108390770B (zh) 一种信息生成方法、装置及服务器
US11310315B1 (en) Techniques for directive-based messaging synchronization
CN108966160B (zh) 一种短信处理方法、装置及计算机可读存储介质
US10402226B2 (en) Media analysis and processing framework on a resource restricted device
US9338241B2 (en) Tracking contacts across multiple communications services
US11704205B2 (en) Systems and methods for transitioning from legacy computer systems
US11722443B1 (en) Techniques for media item display configuration
CN112882886A (zh) 一种软件使用时长的统计方法和装置
CN111917633A (zh) 消息发送方法、发送装置、电子设备及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17830480

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2019503319

Country of ref document: JP

Kind code of ref document: A

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 20197005197

Country of ref document: KR

Kind code of ref document: A

ENP Entry into the national phase

Ref document number: 2017830480

Country of ref document: EP

Effective date: 20190222