WO2018130220A1 - 消息推送方法、设备及可编程设备 - Google Patents

消息推送方法、设备及可编程设备 Download PDF

Info

Publication number
WO2018130220A1
WO2018130220A1 PCT/CN2018/072638 CN2018072638W WO2018130220A1 WO 2018130220 A1 WO2018130220 A1 WO 2018130220A1 CN 2018072638 W CN2018072638 W CN 2018072638W WO 2018130220 A1 WO2018130220 A1 WO 2018130220A1
Authority
WO
WIPO (PCT)
Prior art keywords
push
user
message
target
pushing
Prior art date
Application number
PCT/CN2018/072638
Other languages
English (en)
French (fr)
Inventor
易凌云
Original Assignee
广州市动景计算机科技有限公司
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 广州市动景计算机科技有限公司 filed Critical 广州市动景计算机科技有限公司
Publication of WO2018130220A1 publication Critical patent/WO2018130220A1/zh

Links

Images

Classifications

    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • 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
    • G06F16/951—Indexing; Web crawling techniques
    • 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
    • H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources

Definitions

  • the present invention relates to the field of data push technology, and in particular, to a message push method, device, and programmable device.
  • the message push is usually carried out by an information dissemination platform (for example, an application service platform with a message push function), and the information that the user expects to obtain (such as news, advertisements, etc.) is pushed to the user in the form of a message, which can effectively improve the efficiency of information release.
  • the message push is usually performed by estimating the user's interest to obtain a corresponding message. For example, after estimating the interest of the target user, the message predicted by the user having the same interest feature is pushed to the target that has not read the message.
  • the estimated user interest may not accurately reflect the individual behavior of the user browsing the message, and the corresponding push message actually It is not always possible to get the expected click conversion rate (the probability that a pushed message is clicked or viewed by a user), and the efficiency of message push is not high.
  • a message pushing method comprising:
  • the step of determining the target user of the target message to be pushed includes:
  • a user that matches the user feature with a message feature of the target message is determined to be the target user.
  • the user feature includes at least one of historical behavior information, activity information, interest classification label, location information, and device information of the user;
  • the message feature includes at least message classification information, keyword information, and quality.
  • One of the evaluation information is
  • the first push user is the target user that meets a predetermined activity condition
  • the first push result includes at least a push click conversion rate, and the first push click conversion rate is obtained according to the browsing data of the target message by the first push user;
  • the target message is pushed to the target user other than the pre-push user, otherwise, the target user other than the pre-push user is no longer pushed.
  • the target message When the pre-push click conversion rate satisfies a predetermined threshold, the target message is pushed to the target user other than the pre-push user, otherwise, the target user other than the pre-push user is no longer pushed. The target message.
  • the target message includes a primary message and at least one associated message associated with the primary message;
  • the step of selecting the number of the first-push users that meet the pre-rendering ratio from the plurality of target users to push the target message to obtain the corresponding pre-push result includes:
  • the pre-push result including at least a packet pre-push click conversion rate of each of the pre-push user packets
  • the group first push click conversion rate is obtained according to the browsing data of the corresponding pushed message according to the target user included in the first push user group;
  • the step of pushing the target message to the target user other than the preemptive user according to the obtained pre-push result is:
  • a message pushing device including:
  • a target determining unit configured to determine a plurality of target users of the target message to be pushed
  • a first push unit configured to select, from the plurality of target users, a number of pre-pushing users that meet a pre-rendering ratio, to push the target message to obtain a corresponding pre-push result
  • a pushing unit configured to push the target message to the target user other than the pre-emptive user according to the obtained pre-push result.
  • the target determining unit is further configured to:
  • a user that matches the user feature with a message feature of the target message is determined to be the target user.
  • the first push user is the target user that meets a predetermined activity condition
  • the first push result includes at least a push click conversion rate, and the first push click conversion rate is obtained according to the browsing data of the target message by the first push user;
  • the push unit is further configured to:
  • the target message is pushed to the target user other than the pre-push user, otherwise, the target user other than the pre-push user is no longer pushed.
  • the target message When the pre-push click conversion rate satisfies a predetermined threshold, the target message is pushed to the target user other than the pre-push user, otherwise, the target user other than the pre-push user is no longer pushed. The target message.
  • the target message includes a primary message and at least one associated message associated with the primary message;
  • the first push unit is further configured to:
  • the pre-push result including at least a packet pre-push click conversion rate of each of the pre-push user packets
  • the group first push click conversion rate is obtained according to the browsing data of the corresponding pushed message according to the target user included in the first push user group;
  • the push unit is further configured to:
  • a programmable device comprising a memory and a processor for storing instructions for controlling the processor to operate to perform the first aspect of the invention Any of the message push methods.
  • the first push result is obtained by selecting the first push user from the target user to be pushed the target message, and the target message is pushed to the target user other than the first push user according to the first push result, thereby ensuring the acquisition.
  • Expected message push effect improve message push efficiency.
  • FIG. 1 is a block diagram showing an example of a hardware configuration that can be used to implement an implementation environment of an embodiment of the present invention.
  • FIG. 2 is a flow chart showing a message pushing method according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of an example of a message pushing method according to an embodiment of the present invention.
  • FIG. 4 shows a block diagram of a message push device of an embodiment of the present invention.
  • FIG. 1 is a block diagram showing a hardware configuration of an implementation environment 1000 in which an embodiment of the present invention can be implemented.
  • the implementation environment 1000 includes a server 1100, a client 1200, and a network 1300.
  • the server 1100 can be, for example, a blade server or the like.
  • server 1100 can be a computer.
  • the server 1100 can include a processor 1110, a memory 1120, an interface device 1130, a communication device 1140, a display device 1150, and an input device 1160, as shown in FIG.
  • the server may also include a speaker, a microphone, etc., these components are not relevant to the present invention and are therefore omitted herein.
  • the processor 1110 may be, for example, a central processing unit CPU, a microprocessor MCU, or the like.
  • the memory 1120 includes, for example, a ROM (Read Only Memory), a RAM (Random Access Memory), a nonvolatile memory such as a hard disk, and the like.
  • the interface device 1130 includes, for example, a USB interface, a serial interface, and the like.
  • the communication device 1140 can perform, for example, wired or wireless communication.
  • the display device 1150 is, for example, a liquid crystal display.
  • Input device 1160 can include, for example, a touch screen, a keyboard, and the like.
  • the client 1200 can be a portable computer (1200-1), a desktop computer (1200-2), a mobile phone (1200-3), a tablet (1200-4), and the like.
  • the communication network 1300 can be a wireless network or a network, and can be a local area network or a wide area network.
  • the clients 1200-1, 1200-2, 1200-3, 1200-4, and the server 1100 can communicate over the communication network 1300.
  • the configuration environment 1100 shown in Figure 1 is merely illustrative and is in no way intended to limit the invention, its application or use.
  • the memory 1120 of the server 1100 is configured to store an instruction for controlling the processor 1110 to perform any of the message pushing methods provided by the embodiments of the present invention.
  • server 1100 may relate only to some of the devices therein, for example, server 1100 only relates to processor 1110 and storage device 1120, and the like.
  • a technician can design instructions in accordance with the disclosed aspects of the present invention. How the instructions control the processor for operation is well known in the art and will not be described in detail herein.
  • a message recommendation method is provided in this embodiment. As shown in FIG. 2, the method includes: step S2100 to step S2300.
  • step S2100 a plurality of target users to be pushed target messages are determined.
  • the target message is usually an information dissemination platform (such as a browser, an application service platform with a browser function, a message push platform, etc.) that is expected to be displayed by the Internet to the user who uses the corresponding client device to trigger the user to browse information, such as news. , advertising, Weibo, novels, etc.
  • the target message is news, and the news that meets the requirements of the heat or the timeliness can be obtained as a target message through the method of web crawling and news clustering, and is not enumerated here.
  • the target user is an object to be pushed to the target message
  • the corresponding user is usually determined as the target user after the content of the target message matches the browsing interest point of the user, but in actual application, the content of the message is The user's own interest points are matched, and the matching range is relatively simple.
  • the target click message cannot be obtained to obtain a higher click conversion rate (the click conversion rate is the probability that the user clicks or browses the message after the message is pushed).
  • the message pushing method provided in this embodiment further includes: extracting a user feature of the user; and determining, by the user feature, a user that matches a message feature of the target message as the target user.
  • the user feature is representative information related to the browsing information of the user.
  • the feature of the user includes at least one of historical behavior information, activity information, interest classification label, location information, and device information of the user.
  • the historical behavior information may include browsing history, push history (received push message), click history (push message that has been opened), keyword information (browsing keywords, weights corresponding to keywords, etc.);
  • the activity information may include an active time (which may be accurate to an hour level, such as 9-10 am), an activity level (daily/week/month multiple active); and the interest classification label is a classification label of the user's browsing interest.
  • the device information may include software of a device used by the user Information (software version information dissemination platform or receive push message client version, etc.) and hardware information (device model, operating system information equipment, network equipment, etc.).
  • a message feature is representative information associated with a message.
  • the message feature is at least one of message classification information, keyword information, and quality evaluation information, where the message classification information may be a classification topic of the message, such as sports, entertainment, funny, etc., or Further subdivided vertical classes, such as sports, may be subdivided into soccer, basketball, tennis, etc.;
  • the keyword information may be a series of keywords that may be representative of the message content based on the message content, and may further include Corresponding keyword weights;
  • the quality evaluation information may include a historical click conversion rate of the message, a historical user evaluation of the message, and a quality score obtained by evaluating the quality content model based on the message content.
  • the user feature is evaluated by a preset user evaluation model to obtain a user score, and the user score may be integrated into one dimension or multiple dimensions.
  • the message feature of the target message is evaluated by the corresponding preset message evaluation model to obtain a message score, wherein the user evaluation model and the message evaluation model may be preset according to a specific application scenario, and the message of the target message is used.
  • the difference between the score and the user score is within a predetermined threshold (the threshold may be set according to the engineering experience value or the simulation experiment value), the user characteristic of the user matches the message feature of the target message, and the user is the target message to be pushed.
  • the threshold may be set according to the engineering experience value or the simulation experiment value
  • a message library can be constructed, the message library contains all the target messages to be pushed, and a push message list is maintained for each user, and the message list contains the message features and corresponding user features contained in the message library.
  • the matching message when a certain target message in the message library is to be pushed, may traverse the message push list of all users, and determine the corresponding user in the message push list that includes the target message as the target user.
  • the message pushed to the user can be determined from multiple dimensions, and the user's demand can be more accurately matched to improve the message push. After the actual click conversion rate.
  • Step S2200 Select, from the plurality of target users, a number of pre-pushing users that meet the pre-rendering ratio, and push the target message to obtain a corresponding pre-push result.
  • the first push user is a user that pushes a target message from a plurality of target users; the number of the first push users is in accordance with a first push ratio, and the first push ratio may be set according to an actual application scenario, and may be a corresponding project.
  • the empirical value or the experimental simulation value is usually pushed by a ratio of not more than 50% to ensure that the number of first-pushing users is a minority among the total number of target users, so that the first-pushing result of the corresponding acquisition can reflect the real-time push result of the target message. It does not affect the push results of the target message to most target users.
  • the first push user may be a randomly selected number of users in the target user, for example, the first push ratio is 40%, the target user total number is 100, and 40 target users may be randomly selected as the first.
  • Push users may be a randomly selected number of users in the target user, for example, the first push ratio is 40%, the target user total number is 100, and 40 target users may be randomly selected as the first. Push users.
  • the first push user can also be a user who meets a predetermined selection rule.
  • the first push user may be a user who satisfies a predetermined activity condition, and the predetermined activity condition is used to filter a user with high activity as a pre-push user, for example, the predetermined activity condition is that the user's activity evaluation is in descending order.
  • the predetermined sorting range corresponds to the pre-rendering ratio.
  • a user who satisfies a predetermined activeness condition has a first push ratio of 40%, and a corresponding predetermined sorting range is 1-40, and the user of the 100 target users After the activity evaluation is sorted in descending order, the target users ranked at 1-40 are used as the first push users satisfying the predetermined activity level, wherein the activity evaluation may be based on the active time of the user, the active period, and the activity of the user similar to the user (active) Comprehensive evaluation of time and active period), number of users similar to users, etc.
  • the user with high activity is used as the first push user, and the pre-push result after the target message is pushed can be obtained more quickly, for example, the click conversion rate of the target message is first pushed by the user.
  • the first push result includes at least a push click conversion rate, and the first push click conversion rate is determined according to the first push user
  • the target message may further include a primary message and at least one associated message associated with the primary message, where the associated message may be a similar message having a certain degree of similarity to the message of the primary message, for example, may be
  • the related message belongs to the same topic as the main message, and the associated message may also be the recall message of the main message.
  • the main message is entertainment news n1
  • the user who has browsed the entertainment news n1 also browses the funny class.
  • News n3 then n3 is the recall message of n1.
  • the target message includes a main message and an associated message
  • the main message and the associated message have a certain degree of similarity, corresponding to the same target user to be pushed, and may push the main message to the partial push user first, and the partial push The user pushes the associated message, respectively obtains the corresponding pre-push result, and improves the pre-push efficiency of the pre-push target message to obtain the pre-push result.
  • step S2200 of the message pushing method in the above example may include:
  • the pre-push result including at least a packet pre-push click conversion rate of each of the pre-push user packets
  • the group first push click conversion rate is obtained according to the browsing data of the corresponding pushed message according to the target user included in the first push user group.
  • the pre-pushing user into a predetermined number of pre-pushing user packets, so that the main message or the associated message is pushed only for each of the pre-push user packets, the predetermined number and the number of messages included in the target message. For example, if the target message includes a main message and an associated message, the predetermined number is 2, and the first push user is a group of 2 push user groups, wherein one push user packet only pushes the main message, and the other pushes the packet only. The associated message is pushed. When the target message includes the primary message and the multiple associated messages, the number of first pushed user packets also increases. In addition, the number of users included in each pre-populated user group may be the same, for example, the number of users pushing the user is 20, and the number of users per group pushing the user group is 10.
  • the pre-push result includes at least a packet pre-push conversion rate of each of the pre-push user packets, and the packet first push click conversion rate is correspondingly pushed according to the target user pair included in the pre-push user packet.
  • the browsing data of the message is obtained.
  • the target message includes entertainment news n1 (main message) and funny news n3 (association message), and 20 target users are selected as the first push user among the 100 target users, and the user is first pushed to push the user group as the first group.
  • Each of the first push user groups includes 10 target users, and only pushes n1 for one push user group, n3 for another first push user group, and obtains browse data of corresponding first push user group pair n1 and corresponding
  • After the browsing data of the target user pair n3 in the user group is pushed first, specifically, which of the 10 corresponding target users clicks or browses n1, and which of the 10 corresponding target users clicks or browses n3, correspondingly
  • Step S2300 Push the target message to the target user other than the first push user according to the obtained first push result.
  • the first push result is a real-time feedback result obtained after the target message is pushed to the first push user, and the target message is pushed by the target user other than the first push user according to the first push result, and the target message can be obtained better.
  • the message push effect improves the push efficiency of the target message.
  • the first push user is the target user that meets the predetermined activity condition
  • the first push result includes at least a push click conversion rate
  • the first push click conversion rate is based on the first push user to the target
  • the browsing data of the message is obtained.
  • the target message is pushed to the target user other than the first push user, otherwise, no longer The target user other than the push user pushes the target message.
  • the predetermined threshold is that the target message that the push effect meets the expectation should have a click conversion rate, and may be set according to the engineering experience value or the experimental simulation value corresponding to the specific application scenario.
  • the first click conversion rate of the target message satisfies the predetermined threshold, it means that the real-time push effect of the target message satisfies the expectation, and the target message is pushed to the target user other than the first push user, thereby ensuring the expected message push effect and improving the message push.
  • Efficiency otherwise, means that the real-time push effect of the target message does not meet the expectations, so the target message is no longer pushed to the target user other than the first push user, and the user experience is not affected.
  • the target users to be pushed are determined to be pushed by 20%, so 20 preemptive users that meet the activity condition are selected from the 100 target users, and the 20 push users are selected.
  • the push result obtained after pushing the target message n2 is that 19 users browse the target message, so the push conversion rate of the first click is 95%, which is higher than the predetermined threshold by 90%, which means that the real-time push effect of n2 satisfies the expectation, first Push the remaining 80 target users outside the user to push n2, so as to ensure the expected message push effect and improve the message push efficiency. Otherwise, if the click conversion rate is less than 90%, it is no longer other than the first push user. The remaining 80 target users push n2 to avoid affecting the user experience.
  • the target message includes a primary message and at least one associated message associated with the primary message;
  • the pre-push user is divided into a predetermined number of pre-push user packets such that each of the pre-push The user packet only pushes the main message or the associated message, the predetermined number is the same as the number of messages included in the target message;
  • the pre-push result includes at least a packet pre-click click conversion rate of each of the pre-push user packets
  • the group first push click conversion rate is obtained according to the browsing data of the corresponding pushed message of the target user included in the first push user group; correspondingly, step S2300 is: first, the target message and the optimal grouping Pushing a message corresponding to the click conversion rate, and pushing the message to the target user other than the first push user.
  • the packet first push click conversion rate is optimal, which means that the corresponding message is optimally pushed in all the messages included in the target message, and the push target is pushed to the target user other than the first push user, so that the push effect is in line with expectations.
  • Improve message delivery efficiency is optimal, which means that the corresponding message is optimally pushed in all the messages included in the target message, and the push target is pushed to the target user other than the first push user, so that the push effect is in line with expectations.
  • the target message includes the main message n1 and the associated message n3. Since there is similarity between n1 and n3, it is determined that the corresponding 100 target users to be pushed are the same, and the first push ratio is 20%, so that 100 target users are selected.
  • 20 pre-push users in accordance with the activity condition and divide 20 pre-push users into 2 push user groups: group 1, group 2, group 1 and group 2 respectively contain 10 target users, and only push n1 for group 1.
  • the message push method provided in this embodiment will be further described below with reference to FIG. 3 as an example.
  • the target message to be pushed is news
  • the device implementing the message pushing method in this example is a news dissemination platform having a function of publishing, pushing, and searching for news.
  • the specific message pushing method includes:
  • step S301 the message library including the target message with the push is continuously constructed.
  • the target message to be pushed is news, and the hot or time-sensitive requirement can be obtained through continuous real-time web crawling and news clustering.
  • the news is added to the message library as the target message, so that the construction of the message library continues;
  • Step S302 Construct a corresponding message push list for each user who uses the news dissemination platform client device, where the message list includes a message feature included in the message library and a message matching the corresponding user feature.
  • user A user characteristics are: entertainment news heavy users, sports news light users, prefer to open the client every day at 10 am, first-tier cities
  • user B user characteristics are: sports news heavy users, preferences The client is opened every day at 5 pm
  • the target message library includes news n1 (message feature is entertainment), news n2 (message feature is sports), and n3 (message feature is funny), among which users who have seen n1 are found. I have read the news n3.
  • the user N selects the message n1 and the associated message n3 of the n1 (recall news) from the message library to match the user characteristics of the user A, and joins the push message list [n1, n3] of A, Similarly, news n2 joins user B's push message list [n2].
  • news n4 messages feature is soccer
  • user B when news n4 (message feature is soccer) is added to the message library, it does not match user A's user feature, but with user B.
  • User feature matching can be added to B's push message list [n4, n2], and so on, can achieve continuous construction of a message push list for each user;
  • Step S303 determining the push period of the push target message, if yes, proceeding to step S304, returning to step S301, wherein the push period may be set according to the engineering experience value or the experimental simulation value, and a target message may be pushed every cycle, or Continuously pushing a plurality of target messages, which is not limited in this example, and the target message pushed in the push period will be deleted from the corresponding user push message list;
  • Step S304 determining a plurality of target users to be pushed to the target message, in this example, traversing the user's message list, and determining the target user by the user who includes the target message in the message list;
  • Step S305 it is determined whether the main message and the associated message are included in the target message, for example, in step S302, the news n1 is the main message, and the recall news n3 is the associated message, if not, the process proceeds to step S306, if yes, the process proceeds to step S311;
  • Step S306 the user is selected from the number of target users according to the pre-rendering ratio and the activeness condition is satisfied.
  • the target message is n2
  • the first push ratio is set to 20%, the target users who are ranked in the top 20 are selected from the 100 target users as the first push users;
  • Step S307 the push target message is pushed to the user, and the push conversion rate is obtained according to the browsing data of the target message by the user;
  • Step S308 determining whether the first push click conversion rate meets a predetermined threshold, if yes, proceeds to step S309, otherwise proceeds to step S310;
  • step S307 For example, if the predetermined threshold is 90%, and the pre-push click conversion rate obtained in step S307 is 75%, the process proceeds to step S310;
  • Step S309 pushing the target message to the target user other than the first push user
  • Step S310 the target message is not pushed to the target user other than the first push user
  • the remaining 80 users of the 100 target users do not push n2, and then return to step S301;
  • Step S311 selecting a number of pre-pushing users that meet a predetermined ratio and dividing into a predetermined number of pre-push user packets, the predetermined number being the same as the number of messages included in the target message;
  • the target messages are n1 and n3, and it is determined in step S304 that there are 100 target users to be pushed, and the first push ratio is set to 20%, then 20 users are randomly selected from 100 target users as the first push users, and divided.
  • packet 1 and packet 2 For two pre-push user groups: packet 1 and packet 2, packet 1 and packet 2 respectively contain 10 target users;
  • Step S312 pushing only the main message or the associated message for each pre-pushing user packet to obtain a packet pre-push conversion rate of each of the pre-push user packets;
  • Step S313 the message corresponding to the optimal packet first push click conversion rate in the target message is pushed to the target user other than the first push user;
  • step S312 the packet corresponding to the n3 push-pull click conversion rate is optimal, so the association message n3 is pushed to the remaining 80 target users among the 100 target users.
  • the message pushing method provided in this embodiment has been described above with reference to the accompanying drawings and examples.
  • the first push result is obtained by pushing the user from the target user to be pushed the target message to push the target message, and the first push result is based on the first push result.
  • the target user other than the user pushes the target message, which can ensure the expected message push effect and improve the message push efficiency.
  • a message pushing device 4000 is further provided. As shown in FIG. 4, the method includes: a target determining unit 4100, a pre-pushing unit 4200, and a pushing unit 4300, which are used to implement the message pushing method provided in this embodiment. This will not be repeated here.
  • the message pushing device 4000 includes:
  • a target determining unit 4100 configured to determine a plurality of target users to be pushed by the target message
  • a first push unit configured to select, from the plurality of target users, a number of pre-pushing users that meet a pre-rendering ratio, to push the target message to obtain a corresponding pre-push result
  • the pushing unit 4200 is configured to: according to the obtained first push result, the target user 4300 other than the first push user, push the target message.
  • the target determining unit 4100 is further configured to:
  • a user that matches the user feature with a message feature of the target message is determined to be the target user.
  • the first push user is the target user that meets a predetermined activity condition
  • the first push result includes at least a push click conversion rate, and the first push click conversion rate is obtained according to the browsing data of the target message by the first push user;
  • the pushing unit 4300 is further configured to:
  • the target message is pushed to the target user other than the pre-push user, otherwise, the target user other than the pre-push user is no longer pushed.
  • the target message When the pre-push click conversion rate satisfies a predetermined threshold, the target message is pushed to the target user other than the pre-push user, otherwise, the target user other than the pre-push user is no longer pushed. The target message.
  • the target message includes a primary message and at least one associated message associated with the primary message;
  • the pre-push unit 4200 is further configured to:
  • the pre-push result including at least a packet pre-push click conversion rate of each of the pre-push user packets
  • the group first push click conversion rate is obtained according to the browsing data of the corresponding pushed message according to the target user included in the first push user group;
  • the pushing unit 4300 is further configured to:
  • a programmable device comprising a memory and a processor, the memory for storing an instruction for controlling the processor to perform any one of the embodiments provided in this embodiment Item push method.
  • the physical form of the programmable device can be server 1100 as shown in FIG.
  • a message pushing method, a device, and a programmable device are provided.
  • the first user is pushed from the target user to be pushed by the target message to push the target message to obtain the first push.
  • the target message is pushed to the target user other than the first push user, which can ensure the expected message push effect and improve the message push efficiency.
  • the message push device 4000 can be implemented in a variety of ways.
  • the message push device 4000 can be implemented by an instruction configuration processor.
  • the instructions can be stored in the ROM, and when the device is booted, the instructions are read from the ROM into the programmable device to implement the message push device 4000.
  • message push device 4000 can be cured into a dedicated device (eg, an ASIC).
  • the message push device 4000 can be divided into mutually independent units, or they can be implemented together.
  • the message pushing device 4000 may be implemented by one of the various implementations described above, or may be implemented by a combination of two or more of the various implementations described above.
  • the invention can be a system, method and/or computer program product.
  • the computer program product can comprise a computer readable storage medium having computer readable program instructions embodied thereon for causing a processor to implement various aspects of the present invention.
  • the computer readable storage medium can be a tangible device that can hold and store the instructions used by the instruction execution device.
  • the computer readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing.
  • Non-exhaustive list of computer readable storage media include: portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM) Or flash memory), static random access memory (SRAM), portable compact disk read only memory (CD-ROM), digital versatile disk (DVD), memory stick, floppy disk, mechanical encoding device, for example, with instructions stored thereon A raised structure in the hole card or groove, and any suitable combination of the above.
  • a computer readable storage medium as used herein is not to be interpreted as a transient signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (eg, a light pulse through a fiber optic cable), or through a wire The electrical signal transmitted.
  • the computer readable program instructions described herein can be downloaded from a computer readable storage medium to various computing/processing devices or downloaded to an external computer or external storage device over a network, such as the Internet, a local area network, a wide area network, and/or a wireless network.
  • the network may include copper transmission cables, fiber optic transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and/or edge servers.
  • a network adapter card or network interface in each computing/processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium in each computing/processing device .
  • Computer program instructions for performing the operations of the present invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine related instructions, microcode, firmware instructions, state setting data, or in one or more programming languages.
  • the computer readable program instructions can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer, partly on the remote computer, or entirely on the remote computer or server. carried out.
  • the remote computer can be connected to the user's computer through any kind of network, including a local area network (LAN) or wide area network (WAN), or can be connected to an external computer (eg, using an Internet service provider to access the Internet) connection).
  • the customized electronic circuit such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), can be customized by utilizing state information of computer readable program instructions.
  • Computer readable program instructions are executed to implement various aspects of the present invention.
  • the computer readable program instructions can be provided to a general purpose computer, a special purpose computer, or a processor of other programmable data processing apparatus to produce a machine such that when executed by a processor of a computer or other programmable data processing apparatus Means for implementing the functions/acts specified in one or more of the blocks of the flowcharts and/or block diagrams.
  • the computer readable program instructions can also be stored in a computer readable storage medium that causes the computer, programmable data processing device, and/or other device to operate in a particular manner, such that the computer readable medium storing the instructions includes An article of manufacture that includes instructions for implementing various aspects of the functions/acts recited in one or more of the flowcharts.
  • the computer readable program instructions can also be loaded onto a computer, other programmable data processing device, or other device to perform a series of operational steps on a computer, other programmable data processing device or other device to produce a computer-implemented process.
  • instructions executed on a computer, other programmable data processing apparatus, or other device implement the functions/acts recited in one or more of the flowcharts and/or block diagrams.
  • each block in the flowchart or block diagram can represent a module, a program segment, or a portion of an instruction that includes one or more components for implementing the specified logical functions.
  • Executable instructions can also occur in a different order than those illustrated in the drawings. For example, two consecutive blocks may be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending upon the functionality involved.
  • each block of the block diagrams and/or flowcharts, and combinations of blocks in the block diagrams and/or flowcharts can be implemented in a dedicated hardware-based system that performs the specified function or function. Or it can be implemented by a combination of dedicated hardware and computer instructions. It is well known to those skilled in the art that implementation by hardware, implementation by software, and implementation by a combination of software and hardware are equivalent.

Landscapes

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

Abstract

本发明公开了一种消息推送方法、设备及可编程设备。该方法包括：确定待推送目标消息的若干个目标用户；从若干个目标用户中选取数目符合先推比例的先推用户，以推送目标消息获取对应的先推结果；根据所获取的先推结果，对所述先推用户之外的所述目标用户推送所述目标消息。根据本发明实施例，可以确保获取预期的消息推送结果，提高消息推送效率。

Description

消息推送方法、设备及可编程设备 技术领域
本发明涉及数据推送技术领域，更具体地，涉及一种消息推送方法、设备及可编程设备。
背景技术
随着移动互联网技术的飞速发展以及电子设备智能化的普及，人们早已习惯通过访问互联网以获取信息，互联网也成为当今主流的信息传播平台之一，同时，消息推送也成为通过互联网进行信息传播的手段之一。
消息推送是通常由信息传播平台(例如具有消息推送功能的应用服务平台等)将预估用户期望获取的信息(例如新闻、广告等)以消息的形式推送给用户，可以有效提高信息发布的效率。目前消息推送通常是通过预估用户的兴趣获取对应的消息进行推送，例如，预估目标用户的兴趣后，将预估具有同样兴趣特征的用户浏览过的消息，推送给未阅读该消息的目标用户，但是，在实际应用中，由于用户浏览消息的实时个体行为由于环境因素、心理因素等影响，预估的用户兴趣并不一定能精准地体现用户浏览消息的个体行为，对应推送的消息实际并不一定能获取预期的点击转化率(推送的消息被用户点击或浏览的概率)，消息推送的效率并不高。
发明内容
本发明的一个目的是提供一种用于消息推送的新技术方案。
根据本发明的第一方面，提供了一种消息推送方法，包括：
确定待推送目标消息的若干个目标用户；
从所述若干个目标用户中选取数目符合先推比例的先推用户，以推送所述目标消息获取对应的先推结果；
根据所获取的所述先推结果，对所述先推用户之外的所述目标用户推送所述目标消息。
可选地，所述确定待推送目标消息的目标用户的步骤包括：
抽取用户的用户特征；
将所述用户特征与所述目标消息的消息特征匹配的用户，确定为所述目标用户。
进一步可选地，所述用户特征至少包括用户的历史行为信息、活跃度信息、兴趣分类标签、位置信息、设备信息中其中一种；所述消息特征至少包括消息分类信息、关键词信息、质量评价信息中其中一种。
可选地，所述先推用户为满足预定活跃度条件的所述目标用户；
所述先推结果至少包括先推点击转化率，所述先推点击转化率根据所述先推用户对所述目标消息的浏览数据获取；
所述根据所获取的所述先推结果，对所述先推用户之外的所述目标用户推送所述目标消息的步骤为：
当所述先推点击转化率满足预定阈值时，对所述先推用户之外的所述目标用户推送所述目标消息，否则，不再对所述先推用户之外的所述目标用户推送所述目标消息。
可选地，所述目标消息包括主消息和至少一条与所述主消息关联的关联消息；
从所述若干个目标用户中选取数目符合先推比例的先推用户，以推送所述目标消息获取对应的先推结果的步骤包括：
从所述若干个目标用户中选取数目符合先推比例的先推用户,并将所述先推用户划分为预定数目的先推用户分组，所述预定数目与目标消息中包括的消息数目相同，以及
对每个所述先推用户分组仅推送所述主消息或所述关联消息以获取所述先推结果，所述先推结果至少包括每个所述先推用户分组的分组先推点击转化率，所述分组先推点击转化率根据所述先推用户分组中包含的目标用户对对应推送的消息的浏览数据获取；
所述根据所获取的所述先推结果，对所述先推用户之外的所述目标用 户推送所述目标消息的步骤为：
将所述目标消息中与最优的分组先推点击转化率对应的消息，推送给所述先推用户之外的所述目标用户。
根据本发明的第二方面，还提送一种消息推送设备，包括：
目标确定单元，用于确定待推送目标消息的若干个目标用户；
先推单元，用于从所述若干个目标用户中选取数目符合先推比例的先推用户，以推送所述目标消息获取对应的先推结果；
推送单元，用于根据所获取的所述先推结果，对所述先推用户之外的所述目标用户推送所述目标消息。
可选地，所述目标确定单元还用于：
抽取用户的用户特征；
将所述用户特征与所述目标消息的消息特征匹配的用户，确定为所述目标用户。
可选地，所述先推用户为满足预定活跃度条件的所述目标用户；
所述先推结果至少包括先推点击转化率，所述先推点击转化率根据所述先推用户对所述目标消息的浏览数据获取；
所述推送单元还用于：
当所述先推点击转化率满足预定阈值时，对所述先推用户之外的所述目标用户推送所述目标消息，否则，不再对所述先推用户之外的所述目标用户推送所述目标消息。
可选地，所述目标消息包括主消息和至少一条与所述主消息关联的关联消息；
所述先推单元还用于：
从所述若干个目标用户中选取数目符合先推比例的先推用户,并将所述先推用户划分为预定数目的先推用户分组，所述预定数目与目标消息中包括的消息数目相同，以及
对每个所述先推用户分组仅推送所述主消息或所述关联消息以获取所述先推结果，所述先推结果至少包括每个所述先推用户分组的分组先推点击转化率，所述分组先推点击转化率根据所述先推用户分组中包含的目 标用户对对应推送的消息的浏览数据获取；
所述推送单元还用于：
将所述目标消息中与最优的分组先推点击转化率对应的消息，推送给所述先推用户之外的所述目标用户。
根据本发明的第三方面，提供一种可编程设备，包括存储器和处理器，所述存储器用于存储指令，所述指令用于控制所述处理器进行操作以执行本发明的第一方面提供的任意一项消息推送方法。
根据本发明实施例，通过从待推送目标消息的目标用户中选取先推用户以推送目标消息获取先推结果，并根据先推结果对先推用户之外的目标用户推送目标消息，可以确保获取预期的消息推送效果，提高消息推送效率。
通过以下参照附图对本发明的示例性实施例的详细描述，本发明的其它特征及其优点将会变得清楚。
附图说明
被结合在说明书中并构成说明书的一部分的附图示出了本发明的实施例，并且连同其说明一起用于解释本发明的原理。
图1是显示可用于实现本发明实施例的实施环境的硬件配置的例子的框图。
图2示出了本发明实施例的消息推送方法的流程图。
图3是本发明实施例的消息推送方法的例子的示意图。
图4示出了本发明实施例的消息推送设备的框图。
具体实施方式
现在将参照附图来详细描述本发明的各种示例性实施例。应注意到：除非另外具体说明，否则在这些实施例中阐述的部件和步骤的相对布置、数字表达式和数值不限制本发明的范围。
以下对至少一个示例性实施例的描述实际上仅仅是说明性的，决不作为对本发明及其应用或使用的任何限制。
对于相关领域普通技术人员已知的技术、方法和设备可能不作详细讨论，但在适当情况下，所述技术、方法和设备应当被视为说明书的一部分。
在这里示出和讨论的所有例子中，任何具体值应被解释为仅仅是示例性的，而不是作为限制。因此，示例性实施例的其它例子可以具有不同的值。
应注意到：相似的标号和字母在下面的附图中表示类似项，因此，一旦某一项在一个附图中被定义，则在随后的附图中不需要对其进行进一步讨论。
<硬件配置>
图1是示出可以实现本发明的实施例的实施环境1000的硬件配置的框图。如图1所示，实施环境1000包括服务器1100、客户端1200以及网络1300。
服务器1100例如可以是刀片服务器等。在一个例子中，服务器1100可以是一台计算机。在在另一个例子中，服务器1100可以如图1所示，包括处理器1110、存储器1120、接口装置1130、通信装置1140、显示装置1150、输入装置1160。尽管服务器也可以包括扬声器、麦克风等等，但是，这些部件与本发明无关，故在此省略。其中，处理器1110例如可以是中央处理器CPU、微处理器MCU等。存储器1120例如包括ROM(只读存储器)、RAM(随机存取存储器)、诸如硬盘的非易失性存储器等。接口装置1130例如包括USB接口、串行接口等。通信装置1140例如能够进行有线或无线通信。显示装置1150例如是液晶显示屏。输入装置1160例如可以包括触摸屏、键盘等。
客户端1200可以是便携式电脑(1200-1)、台式计算机(1200-2)、手机(1200-3)、平板电脑(1200-4)等。
通信网络1300可以是无线网络也可以网络，可以是局域网也可以是广域网。在图1所示的配置环境1000中，客户端1200-1、1200-2、1200-3、1200-4以及服务器1100可以通过通信网络1300进行通信。
图1所示的配置环境1100仅是解释性的，并且决不是为了要限制本发明、其应用或用途。应用于本发明的实施例中，服务器1100的所述存储 器1120用于存储指令，所述指令用于控制所述处理器1110进行操作以执行本发明实施例提供的任意一项消息推送方法。本领域技术人员应当理解，尽管在图1中对服务器1100示出了多个装置，但是，本发明可以仅涉及其中的部分装置，例如，服务器1100只涉及处理器1110和存储装置1120等。技术人员可以根据本发明所公开方案设计指令。指令如何控制处理器进行操作，这是本领域公知，故在此不再详细描述。
<实施例>
<方法>
本实施例中提供一种消息推荐方法，如图2所示，包括：步骤S2100至步骤S2300。
步骤S2100，确定待推送目标消息的若干个目标用户。
其中，目标消息通常为信息传播平台(例如浏览器、具有浏览器功能的应用服务平台、消息推送平台等)期望通过互联网推送向使用对应客户端设备的用户展示以触发用户浏览的信息，例如新闻、广告、微博、小说等。具体地，例如目标消息是新闻，可以通过网页爬取、新闻聚类的方法获获取符合热度或时效性要求的新闻作为目标消息进行推送，在此不一一列举。
在本实施例中，目标用户是待推送目标消息的对象，通常是在目标消息的内容与用户的浏览兴趣点匹配后将对应的用户确定为目标用户，但在实际应用中，通过消息内容和用户本身的兴趣点匹配，匹配范围较为单一，在实际应用中，并不能使得目标消息推送后获取较高的点击转化率(点击转化率为消息推送后用户点击或浏览该消息的概率)。
因此，在一个例子中，本实施例中提供的消息推送方法还包括：抽取用户的用户特征；将所述用户特征与所述目标消息的消息特征匹配的用户，确定为所述目标用户。
其中，用户特征是与用户自身浏览信息相关的具有代表性的信息，具体地，所述用户的特征至少包括用户的历史行为信息、活跃度信息、兴趣分类标签、位置信息、设备信息中其中一种，所述历史行为信息可以包括 浏览历史、推送历史(已接收的推送消息)、点击历史(已被打开的推送消息)、关键词信息(浏览的关键词、关键词对应的权重等)；所述活跃度信息可以包括活跃时间(可以精确到小时级别，比如上午9-10点)、活跃程度(每日/周/月若干次活跃)；所述兴趣分类标签是用户浏览兴趣的分类标签，可以包括分类类别(体育、娱乐、搞笑等)以及程度分类(重度、中度、轻度等)等；所述位置信息是用户位置(可以精确到经纬度或者国家/行政区/街道/门牌号或标志性建筑等)或者用户在预定时间段内的移动轨迹等，所述设备信息可以包括用户使用的设备的软件信息(信息传播平台的软件版本或接受推送消息的客户端版本等)以及硬件信息(设备型号、设备的操作系统信息、设备使用的网络等)。
消息特征是与消息相关的具有代表性的信息。具体地，所述消息特征至少是消息分类信息、关键词信息、质量评价信息中其中一种，其中，消息分类信息可以是消息的分类主题，例如体育类、娱乐类、搞笑类等，还可是进一步细分的垂类，例如体育类可以细分为足球类、篮球类、网球类等等；关键词信息可以是根据消息内容提取的一系列可以代表消息内容特征的关键词，进一步还可以包括对应的关键词权重；质量评价信息可以包括消息的历史点击转化率、消息的历史用户评价、以及根据消息内容利用质量模型进行评价得到的质量评分等。
在本实施例中，可以对每一个抽取用户特征后，通过预设的用户评价模型对用户特征进行评价以获取用户评分，该用户评分可以是综合一个维度的也可以是多个维度的，同时通过对应的预设的消息评价模型对目标消息的消息特征进行评价以获取消息评分，其中，所述用户评价模型、消息评价模型可以是根据具体的应用场景进行预设，而当目标消息的消息评分与用户评分比较的差值在预定的阈值内时(该阈值可以根据工程经验值或仿真实验值设置)，该用户的用户特征与目标消息的消息特征匹配，该用户为目标消息待推送的目标用户。
更具体地，可以构建一个消息库，消息库中包含待推送的所有目标消息，同时对于每个用户维护一个推送消息列表，该消息列表中包含的是消息库中包含的消息特征与对应用户特征匹配的消息，当消息库中某一条目 标消息待推送时，可以遍历所有用户的消息推送列表，将消息推送列表中包含该目标消息的对应的用户确定为目标用户。
在上述例子中，通过将用户特征与目标消息的消息特征匹配的用户确定为目标用户，可以从多个维度综合匹配确定向用户推送的消息，更精准地贴合用户的需求，以提升消息推送后的实际点击转化率。
步骤S2200，从所述若干个目标用户中选取数目符合先推比例的先推用户，以推送所述目标消息获取对应的先推结果。
其中，所述先推用户是从若干个目标用户中选取的预先推送目标消息的用户；该先推用户的数目符合先推比例，该先推比例可以根据实际应用场景设置，可以是对应的工程经验值或者实验仿真值，通常先推比例不大于50％，以确保先推用户的数目在目标用户的全部数目中占少数，使得对应获取的先推结果既可以反映目标消息的实时推送结果，又不会影响目标消息对大部分目标用户的推送结果。
在本实施例中，先推用户可以是在目标用户中随机选取的数目符合先推比例的用户，例如，先推比例为40％，目标用户总数为100，可以随机选取40个目标用户作为先推用户。
先推用户也可以是符合某个预定选取规则的用户。例如，先推用户可以是满足预定活跃度条件的用户，所述预定活跃度条件用于筛选活跃度高的用户作为先推用户，例如所述预定活跃度条件是用户的活跃度评价降序排序在预定排序范围内，所述预定排序范围与所述先推比例对应，例如，满足预定活跃度条件的用户，先推比例为40％，对应的预定排序范围为1-40,100个目标用户中用户的活跃度评价降序排序后，排序在1-40的目标用户作为满足预定活跃度条件的先推用户，其中活跃度评价可以根据用户的活跃时间、活跃周期、与用户相似的用户的活跃度(活跃时间及活跃周期)、与用户相似的用户的数量等综合评价。
将活跃度高的用户作为先推用户，可以更快地获取目标消息推送后的先推结果，例如先推用户对所述目标消息的点击转化率。对应地，当所述先推用户为满足预定活跃度条件的所述目标用户时，所述先推结果至少包括先推点击转化率，所述先推点击转化率根据所述先推用户对所述目标消 息的浏览数据获取。例如，将活跃度评价降序排序在1-40的前40个目标用户为先推用户，在获取40个目标用户对目标消息的浏览数据，具体地，可以是40个先推用户中哪些用户点击或浏览了目标消息，假设点击或浏览目标消息的用户为20个，就可以获取先推点击转化率为20/40*100％＝50％。
而在另一个例子中，目标消息还可以包括主消息和至少一条与所述主消息关联的关联消息，其中，关联消息可以是与主消息的消息具有一定相似度的相似消息，例如，可以是与主消息属于同一个主题的相关消息，关联消息也可以是主消息的召回消息，例如，主消息是娱乐类新闻n1，在统计后发现浏览过娱乐类新闻n1的用户也都浏览了搞笑类新闻n3，则n3为n1的召回消息。
在上例中，目标消息包括主消息和关联消息，主消息和关联消息之间具有一定的相似度，对应待推送的目标用户相同，可以分别对部分先推用户推送主消息、对部分先推用户推送关联消息，分别获取对应的先推结果，提高预先推送目标消息获取先推结果的先推效率。
因此，在上例中的消息推送方法的步骤S2200可以包括：
从所述若干个目标用户中选取数目符合先推比例的先推用户,并将所述先推用户划分为预定数目的先推用户分组，所述预定数目与目标消息中包括的消息数目相同，以及
对每个所述先推用户分组仅推送所述主消息或所述关联消息以获取所述先推结果，所述先推结果至少包括每个所述先推用户分组的分组先推点击转化率，所述分组先推点击转化率根据所述先推用户分组中包含的目标用户对对应推送的消息的浏览数据获取。
其中，从目标用户中选取先推用户在前文已经描述，在此不再赘述。而将先推用户划分为预定数目的先推用户分组，可以使得对每个所述先推用户分组仅推送所述主消息或所述关联消息，所述预定数目与目标消息中包括的消息数目相同，例如，目标消息包括主消息和一条关联消息，则预定数目为2，所述先推用户为2组先推用户分组，其中一个先推用户分组只推送主消息，另一个先推分组只推送关联消息，当目标消息包括主消息和多条关联消息时，先推用户分组的数目也随之递增。此外，每个先推用 户分组中包含的用户数目可以是相同的，例如，先推用户的用户数目为20，每一组先推用户分组的用户数是10。
对应地，所述先推结果至少包括每个所述先推用户分组的分组先推点击转化率，所述分组先推点击转化率根据所述先推用户分组中包含的目标用户对对应推送的消息的浏览数据获取。
例如，目标消息包括娱乐类新闻n1(主消息)和搞笑类新闻n3(关联消息)，在100个目标用户中选取20个目标用户作为先推用户，先推用户为2组先推用户分组，每个先推用户分组中包括10个目标用户，对一个先推用户分组只推送n1，对另一个先推用户分组只推送n3，在获取对应的先推用户分组对n1的浏览数据和对应的先推用户分组中目标用户对n3的浏览数据后，具体地，可以是10个对应的目标用户中哪些点击或浏览了n1，以及10个对应的目标用户中哪些点击或浏览了n3，对应地，假设浏览了n1的用户数为5，浏览了n3的用户数为9，则n1对应的先推点击转化率为5/10*100％＝50％,n3对应的先推点击转化率9/10*100％＝90％。
步骤S2300，根据所获取的所述先推结果，对所述先推用户之外的所述目标用户推送所述目标消息。
获取先推结果后，该先推结果是目标消息推送给先推用户后获取的实时反馈结果，根据该先推结果对先推用户之外的目标用户进行目标消息的推送，可以获取较好的消息推送效果，提高目标消息的推送效率。
具体地，所述先推用户为满足预定活跃度条件的所述目标用户，所述先推结果至少包括先推点击转化率，所述先推点击转化率根据所述先推用户对所述目标消息的浏览数据获取，对应地，步骤S2300为：当所述先推点击转化率满足预定阈值时，对所述先推用户之外的所述目标用户推送所述目标消息，否则，不再对所述先推用户之外的所述目标用户推送所述目标消息。
其中，该预定阈值是推送效果满足预期的目标消息应该具有点击转化率，可以根据具体应用场景对应的工程经验值或者实验仿真值设置。
当目标消息的先推点击转化率满足预定阈值时，意味着目标消息的实时推送效果满足预期，对先推用户之外的目标用户推送目标消息，可以确 保获取预期的消息推送效果，提高消息推送效率，否则，意味着目标消息的实时推送效果不满足预期，因此不再将目标消息推送给先推用户之外的目标用户，避免影响用户体验。
例如，对于目标消息n2，确定待推送的100个目标用户，先推比例为20％，因此从100个目标用户中选取了符合活跃度条件的20个先推用户，对这20个先推用户推送目标消息n2后获取的先推结果为其中19个用户浏览了该目标消息，因此先推点击转化率为95％，高于预定阈值90％，意味着n2的实时推送效果满足预期，对先推用户之外的剩余80个目标用户推送n2，从而确保获取预期的消息推送效果，提高消息推送效率，否则，若先推点击转化率低于90％，不再对对先推用户之外的剩余80个目标用户推送n2，避免影响用户体验。
或者，具体地，所述目标消息包括主消息和至少一条与所述主消息关联的关联消息；所述先推用户被划分为预定数目的先推用户分组，以使得对每个所述先推用户分组仅推送所述主消息或所述关联消息，所述预定数目与目标消息中包括的消息数目相同；所述先推结果至少包括每个所述先推用户分组的分组先推点击转化率，所述分组先推点击转化率根据所述先推用户分组中包含的目标用户对对应推送的消息的浏览数据获取；对应地，步骤S2300为：将所述目标消息中与最优的分组先推点击转化率对应的消息，推送给所述先推用户之外的所述目标用户。
其中，分组先推点击转化率最优，意味着对应的消息在目标消息包含的所有消息中实时推送效果最优，对先推用户之外的目标用户推送该消息，可以使得推送效果符合预期，提高消息推送效率。
例如，目标消息包括主消息n1和关联消息n3，由于n1和n3之间具有相似度，确定对应待推送的100个目标用户相同，先推比例为20％，因此从100个目标用户中选取了符合活跃度条件的20个先推用户，并将20个先推用户划分为2个推用户分组：分组1、分组2，分组1和分组2分别包含10个目标用户，对分组1只推送n1以及对分组2只推送n3，分组1中有9个用户浏览了n1，则对应的分组先推点击转化率9/10*100＝90％,分组2中有7个用户浏览了n3，则对应的分组先推点击转化率7/10*100＝70％, 推送n1的分组先推点击转化率优于推送n3的分组先推点击转化率，则意味着向目标用户推送n1的效果最好，因此，对先推用户之外的80个目标用户推送n1，类似地，如果推送n3的分组先推点击转化率高于推送n1的分组先推点击转化率，对先推用户之外的80个目标用户推送n3，最终使得大部分目标用户被推送的是推送效果最好的消息，可以使得消息推送效果符合预期，提高消息推送效率。
<例子>
以下将结合图3为例，进一步说明本实施例中提供的消息推送方法。
在本例中，待推送的目标消息为新闻，实施本例中的消息推送方法的设备为具有发布、推送、搜索新闻功能的新闻传播平台，具体的消息推送方法包括：
步骤S301，持续构建包括带推送的目标消息的消息库，在本例中，待推送的目标消息为新闻，可以通过持续实时的网页爬取、新闻聚类的方法获获取符合热度或时效性要求的新闻作为目标消息加入消息库中，使得消息库的构建持续进行；
步骤S302，对每个使用新闻传播平台客户端设备的用户构建对应的消息推送列表，该消息列表中包含的是消息库中包含的消息特征与对应用户特征匹配的消息，
例如，用户A，用户特征为：是娱乐类新闻重度用户，体育类新闻轻度用户，偏好每天上午10点打开客户端，在一线城市；用户B，用户特征为：体育类新闻重度用户，偏好每天下午5点打开客户端；而目标消息库中包括新闻n1(消息特征为娱乐类)、新闻n2(消息特征为体育类)以及n3(消息特征为搞笑类)，其中发现看过n1的用户都看了新闻n3，因此，对用户A从消息库中选取消息特征与用户A的用户特征匹配的消息n1以及n1的关联消息n3(召回新闻)加入A的推送消息列表[n1,n3],类似地，也新闻n2加入用户B的推送消息列表[n2]，此外，当消息库中新增新闻n4(消息特征为足球类)时，与用户A的用户特征并不匹配、但与用户B的用户特征匹配，可以加入B的推送消息列表[n4,n2]，以此类推，可以实现持续对每个用户构建消息推送列表；
步骤S303，判断到达推送目标消息的推送周期，如果是，进入步骤S304，回到步骤S301，其中，推送周期可以根据工程经验值或者实验仿真值设置，每到周期可以推送一条目标消息，也可以连续推送多条目标消息，在本例中并不限制，在推送周期内推送完成的目标消息将从对应的用户推送消息列表中删除；
步骤S304，确定待推送目标消息的若干个目标用户，在本例中，可以遍历用户的消息列表，将消息列表中包含目标消息的用户确定目标用户；
步骤S305，判断目标消息中是否包含主消息和关联消息，例如，在步骤S302中新闻n1为主消息，其召回新闻n3为关联消息，如果不包含，进入步骤S306，如果包含，进入步骤S311；
步骤S306，从所述若干个目标用户中选取数目符合先推比例且满足活跃度条件先推用户，在本例中，假设目标消息为n2，从步骤S304中确定存在100个待推送的目标用户，先推比例设置为20％，则从100个目标用户中选取活跃度评价排序在前20名的目标用户作为先推用户；
步骤S307，对先推用户推送目标消息，根据先推用户对目标消息的浏览数据获取先推点击转化率；
例如，20个先推用户中有15个用户浏览了目标消息n2，则先推点击转化率为15/20*100％＝75％；
步骤S308，判断先推点击转化率是否满足预定阈值，如果满足，进入步骤S309，否则进入步骤S310；
例如，预定阈值为90％，步骤S307中获取的先推点击转化率75％，则进入步骤S310；
步骤S309，对所述先推用户之外的所述目标用户推送所述目标消息；
例如，对100个目标用户中剩余80个用户推送n2，之后回到步骤S301；
步骤S310，不再对所述先推用户之外的所述目标用户推送所述目标消息；
例如，对100个目标用户中剩余80个用户不推送n2，之后回到步骤S301；
步骤S311,选取数目符合预定比例的先推用户并划分为预定数目的 先推用户分组，所述预定数目与目标消息中包括的消息数目相同；
例如，目标消息为n1和n3，从步骤S304中确定存在100个待推送的目标用户，先推比例设置为20％，则从100个目标用户中随机选取20个用户为先推用户，并划分为2个先推用户分组：分组1和分组2，分组1和分组2分别包含10个目标用户；
步骤S312，对每个先推用户分组仅推送主消息或关联消息以获取每个所述先推用户分组的分组先推点击转化率；
例如，对分组1只推送n1以及对分组2只推送n3，分组1中有7个用户浏览了n1，则对应的分组先推点击转化率7/10*100＝70％,分组2中有9个用户浏览了n3，则对应的分组先推点击转化率9/10*100＝90％；
步骤S313，将目标消息中与最优的分组先推点击转化率对应的消息，推送给先推用户之外的所述目标用户；
例如，在步骤S312的例子中，n3对应的分组先推点击转化率最优，因此将关联消息n3推送给100个目标用户中剩余的80个目标用户。
以上已经结合附图和例子说明了本实施例中提供的消息推送方法，通过从待推送目标消息的目标用户中选取先推用户以推送目标消息获取先推结果，并根据先推结果对先推用户之外的目标用户推送目标消息，可以确保获取预期的消息推送效果，提高消息推送效率。
<消息推送设备>
在本实施例中还提供一种消息推送设备4000，如图4所示，包括：目标确定单元4100、先推单元4200以及推送单元4300，用于实施本实施例中提供的消息推送方法，在此不再赘述。
消息推送设备4000，包括：
目标确定单元4100，用于确定待推送目标消息的若干个目标用户；
先推单元，用于从所述若干个目标用户中选取数目符合先推比例的先推用户，以推送所述目标消息获取对应的先推结果；
推送单元4200，用于根据所获取的所述先推结果，所述先推用户之外的所述目标用户4300，推送所述目标消息。
可选地，所述目标确定单元4100还用于：
抽取用户的用户特征；
将所述用户特征与所述目标消息的消息特征匹配的用户，确定为所述目标用户。
可选地，所述先推用户为满足预定活跃度条件的所述目标用户；
所述先推结果至少包括先推点击转化率，所述先推点击转化率根据所述先推用户对所述目标消息的浏览数据获取；
所述推送单元4300还用于：
当所述先推点击转化率满足预定阈值时，对所述先推用户之外的所述目标用户推送所述目标消息，否则，不再对所述先推用户之外的所述目标用户推送所述目标消息。
可选地，所述目标消息包括主消息和至少一条与所述主消息关联的关联消息；
所述先推单元4200还用于：
从所述若干个目标用户中选取数目符合先推比例的先推用户,并将所述先推用户划分为预定数目的先推用户分组，所述预定数目与目标消息中包括的消息数目相同，以及
对每个所述先推用户分组仅推送所述主消息或所述关联消息以获取所述先推结果，所述先推结果至少包括每个所述先推用户分组的分组先推点击转化率，所述分组先推点击转化率根据所述先推用户分组中包含的目标用户对对应推送的消息的浏览数据获取；
所述推送单元4300还用于：
将所述目标消息中与最优的分组先推点击转化率对应的消息，推送给所述先推用户之外的所述目标用户。
<可编程设备>
在本实施例中，还提供一种可编程设备，包括存储器和处理器，所述存储器用于存储指令，所述指令用于控制所述处理器进行操作以执行本实施例中提供的任意一项消息推送方法。在一个例子中，可编程设备的实体形式可以如图1所示的服务器1100。
以上已经结合附图描述了本实施例，根据本实施例，提供一种消息推 送方法、设备以及可编程设备，通过从待推送目标消息的目标用户中选取先推用户以推送目标消息获取先推结果，并根据先推结果对先推用户之外的目标用户推送目标消息，可以确保获取预期的消息推送效果，提高消息推送效率。
本领域技术人员应当明白，可以通过各种方式来实现消息推送设备4000。例如，可以通过指令配置处理器来实现消息推送设备4000。例如，可以将指令存储在ROM中，并且当启动设备时，将指令从ROM读取到可编程器件中来实现消息推送设备4000。例如，可以将消息推送设备4000固化到专用器件(例如ASIC)中。可以将消息推送设备4000分成相互独立的单元，或者可以将它们合并在一起实现。消息推送设备4000可以通过上述各种实现方式中的一种来实现，或者可以通过上述各种实现方式中的两种或更多种方式的组合来实现。
本领域技术人员公知的是，随着诸如大规模集成电路技术的电子信息技术的发展和软件硬件化的趋势，要明确划分计算机系统软、硬件界限已经显得比较困难了。因为，任何操作可以软件来实现，也可以由硬件来实现。任何指令的执行可以由硬件完成，同样也可以由软件来完成。对于某一机器功能采用硬件实现方案还是软件实现方案，取决于价格、速度、可靠性、存储容量、变更周期等非技术性因素。因此，对于电子信息技术领域的普通技术人员来说，更为直接和清楚地描述一个技术方案的方式是描述该方案中的各个操作。在知道所要执行的操作的情况下，本领域技术人员可以基于对所述非技术性因素的考虑直接设计出期望的产品。
本发明可以是系统、方法和/或计算机程序产品。计算机程序产品可以包括计算机可读存储介质，其上载有用于使处理器实现本发明的各个方面的计算机可读程序指令。
计算机可读存储介质可以是可以保持和存储由指令执行设备使用的指令的有形设备。计算机可读存储介质例如可以是――但不限于――电存储设备、磁存储设备、光存储设备、电磁存储设备、半导体存储设备或者上述的任意合适的组合。计算机可读存储介质的更具体的例子(非穷举的列表)包括：便携式计算机盘、硬盘、随机存取存储器(RAM)、只读存储 器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、静态随机存取存储器(SRAM)、便携式压缩盘只读存储器(CD-ROM)、数字多功能盘(DVD)、记忆棒、软盘、机械编码设备、例如其上存储有指令的打孔卡或凹槽内凸起结构、以及上述的任意合适的组合。这里所使用的计算机可读存储介质不被解释为瞬时信号本身，诸如无线电波或者其他自由传播的电磁波、通过波导或其他传输媒介传播的电磁波(例如，通过光纤电缆的光脉冲)、或者通过电线传输的电信号。
这里所描述的计算机可读程序指令可以从计算机可读存储介质下载到各个计算/处理设备，或者通过网络、例如因特网、局域网、广域网和/或无线网下载到外部计算机或外部存储设备。网络可以包括铜传输电缆、光纤传输、无线传输、路由器、防火墙、交换机、网关计算机和/或边缘服务器。每个计算/处理设备中的网络适配卡或者网络接口从网络接收计算机可读程序指令，并转发该计算机可读程序指令，以供存储在各个计算/处理设备中的计算机可读存储介质中。
用于执行本发明操作的计算机程序指令可以是汇编指令、指令集架构(ISA)指令、机器指令、机器相关指令、微代码、固件指令、状态设置数据、或者以一种或多种编程语言的任意组合编写的源代码或目标代码，所述编程语言包括面向对象的编程语言—诸如Smalltalk、C++等，以及常规的过程式编程语言—诸如“C”语言或类似的编程语言。计算机可读程序指令可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中，远程计算机可以通过任意种类的网络—包括局域网(LAN)或广域网(WAN)—连接到用户计算机，或者，可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。在一些实施例中，通过利用计算机可读程序指令的状态信息来个性化定制电子电路，例如可编程逻辑电路、现场可编程门阵列(FPGA)或可编程逻辑阵列(PLA)，该电子电路可以执行计算机可读程序指令，从而实现本发明的各个方面。
这里参照根据本发明实施例的方法、装置(系统)和计算机程序产品 的流程图和/或框图描述了本发明的各个方面。应当理解，流程图和/或框图的每个方框以及流程图和/或框图中各方框的组合，都可以由计算机可读程序指令实现。
这些计算机可读程序指令可以提供给通用计算机、专用计算机或其它可编程数据处理装置的处理器，从而生产出一种机器，使得这些指令在通过计算机或其它可编程数据处理装置的处理器执行时，产生了实现流程图和/或框图中的一个或多个方框中规定的功能/动作的装置。也可以把这些计算机可读程序指令存储在计算机可读存储介质中，这些指令使得计算机、可编程数据处理装置和/或其他设备以特定方式工作，从而，存储有指令的计算机可读介质则包括一个制造品，其包括实现流程图和/或框图中的一个或多个方框中规定的功能/动作的各个方面的指令。
也可以把计算机可读程序指令加载到计算机、其它可编程数据处理装置、或其它设备上，使得在计算机、其它可编程数据处理装置或其它设备上执行一系列操作步骤，以产生计算机实现的过程，从而使得在计算机、其它可编程数据处理装置、或其它设备上执行的指令实现流程图和/或框图中的一个或多个方框中规定的功能/动作。
附图中的流程图和框图显示了根据本发明的多个实施例的系统、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上，流程图或框图中的每个方框可以代表一个模块、程序段或指令的一部分，所述模块、程序段或指令的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。在有些作为替换的实现中，方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如，两个连续的方框实际上可以基本并行地执行，它们有时也可以按相反的顺序执行，这依所涉及的功能而定。也要注意的是，框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合，可以用执行规定的功能或动作的专用的基于硬件的系统来实现，或者可以用专用硬件与计算机指令的组合来实现。对于本领域技术人员来说公知的是，通过硬件方式实现、通过软件方式实现以及通过软件和硬件结合的方式实现都是等价的。
以上已经描述了本发明的各实施例，上述说明是示例性的，并非穷尽 性的，并且也不限于所披露的各实施例。在不偏离所说明的各实施例的范围和精神的情况下，对于本技术领域的普通技术人员来说许多修改和变更都是显而易见的。本文中所用术语的选择，旨在最好地解释各实施例的原理、实际应用或对市场中的技术改进，或者使本技术领域的其它普通技术人员能理解本文披露的各实施例。本发明的范围由所附权利要求来限定。

Claims (10)

  1. 一种消息推送方法，其特征在于，包括：
    确定待推送目标消息的若干个目标用户；
    从所述若干个目标用户中选取数目符合先推比例的先推用户，以推送所述目标消息获取对应的先推结果；
    根据所获取的所述先推结果，对所述先推用户之外的所述目标用户推送所述目标消息。
  2. 根据权利要求1所述的方法，其特征在于，所述确定待推送目标消息的目标用户的步骤包括：
    抽取用户的用户特征；
    将所述用户特征与所述目标消息的消息特征匹配的用户，确定为所述目标用户。
  3. 根据权利要求1或2所述的方法，其特征在于，
    所述用户特征至少包括用户的历史行为信息、活跃度信息、兴趣分类标签、位置信息、设备信息中其中一种；
    所述消息特征至少包括消息分类信息、关键词信息、质量评价信息中其中一种。
  4. 根据权利要求1-3中任意一项所述的方法，其特征在于，
    所述先推用户为满足预定活跃度条件的所述目标用户；
    所述先推结果至少包括先推点击转化率，所述先推点击转化率根据所述先推用户对所述目标消息的浏览数据获取；
    所述根据所获取的所述先推结果，对所述先推用户之外的所述目标用户推送所述目标消息的步骤为：
    当所述先推点击转化率满足预定阈值时，对所述先推用户之外的所述 目标用户推送所述目标消息，否则，不再对所述先推用户之外的所述目标用户推送所述目标消息。
  5. 根据权利要求1-5中任意一项所述的方法，其特征在于，
    所述目标消息包括主消息和至少一条与所述主消息关联的关联消息；
    从所述若干个目标用户中选取数目符合先推比例的先推用户，以推送所述目标消息获取对应的先推结果的步骤包括：
    从所述若干个目标用户中选取数目符合先推比例的先推用户,并将所述先推用户划分为预定数目的先推用户分组，所述预定数目与目标消息中包括的消息数目相同，以及
    对每个所述先推用户分组仅推送所述主消息或所述关联消息以获取所述先推结果，所述先推结果至少包括每个所述先推用户分组的分组先推点击转化率，所述分组先推点击转化率根据所述先推用户分组中包含的目标用户对对应推送的消息的浏览数据获取；
    所述根据所获取的所述先推结果，对所述先推用户之外的所述目标用户推送所述目标消息的步骤为：
    将所述目标消息中与最优的分组先推点击转化率对应的消息，推送给所述先推用户之外的所述目标用户。
  6. 一种消息推送设备，其特征在于，包括：
    目标确定单元，用于确定待推送目标消息的若干个目标用户；
    先推单元，用于从所述若干个目标用户中选取数目符合先推比例的先推用户，以推送所述目标消息获取对应的先推结果；
    推送单元，用于根据所获取的所述先推结果，对所述先推用户之外的所述目标用户推送所述目标消息。
  7. 根据权利要求6所述的设备，其特征在于，所述目标确定单元还用于：
    抽取用户的用户特征；
    将所述用户特征与所述目标消息的消息特征匹配的用户，确定为所述目标用户。
  8. 根据权利要求6或7所述的设备，其特征在于，
    所述先推用户为满足预定活跃度条件的所述目标用户；
    所述先推结果至少包括先推点击转化率，所述先推点击转化率根据所述先推用户对所述目标消息的浏览数据获取；
    所述推送单元还用于：
    当所述先推点击转化率满足预定阈值时，对所述先推用户之外的所述目标用户推送所述目标消息，否则，不再对所述先推用户之外的所述目标用户推送所述目标消息。
  9. 根据权利要求6-8中任意一项所述的设备，其特征在于：
    所述目标消息包括主消息和至少一条与所述主消息关联的关联消息；
    所述先推单元还用于：
    从所述若干个目标用户中选取数目符合先推比例的先推用户,并将所述先推用户划分为预定数目的先推用户分组，所述预定数目与目标消息中包括的消息数目相同，以及
    对每个所述先推用户分组仅推送所述主消息或所述关联消息以获取所述先推结果，所述先推结果至少包括每个所述先推用户分组的分组先推点击转化率，所述分组先推点击转化率根据所述先推用户分组中包含的目标用户对对应推送的消息的浏览数据获取；
    所述推送单元还用于：
    将所述目标消息中与最优的分组先推点击转化率对应的消息，推送给所述先推用户之外的所述目标用户。
  10. 一种可编程设备，其特征在于，包括存储器和处理器，所述存储器用于存储指令，所述指令用于控制所述处理器进行操作以执行根据权利要求1-5中任意一项所述的方法。
PCT/CN2018/072638 2017-01-16 2018-01-15 消息推送方法、设备及可编程设备 WO2018130220A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710028739.7 2017-01-16
CN201710028739.7A CN108259546A (zh) 2017-01-16 2017-01-16 消息推送方法、设备及可编程设备

Publications (1)

Publication Number Publication Date
WO2018130220A1 true WO2018130220A1 (zh) 2018-07-19

Family

ID=62721326

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/072638 WO2018130220A1 (zh) 2017-01-16 2018-01-15 消息推送方法、设备及可编程设备

Country Status (2)

Country Link
CN (1) CN108259546A (zh)
WO (1) WO2018130220A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110413895A (zh) * 2019-07-31 2019-11-05 腾讯科技（深圳）有限公司 一种信息推送的方法以及相关装置
CN111163152A (zh) * 2019-12-26 2020-05-15 中国平安人寿保险股份有限公司 信息推送方法及装置
CN112287236A (zh) * 2020-11-19 2021-01-29 每日互动股份有限公司 文本消息的推送方法、装置、计算机设备及存储介质
CN113312554A (zh) * 2021-06-15 2021-08-27 北京百度网讯科技有限公司 用于评价推荐系统的方法及装置、电子设备和介质

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110875949A (zh) * 2018-09-04 2020-03-10 京东数字科技控股有限公司 用于推送信息的方法和装置
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
CN111131359B (zh) * 2018-10-31 2022-06-07 北京字节跳动网络技术有限公司 用于生成信息的方法和装置
CN111131354B (zh) * 2018-10-31 2022-05-10 北京字节跳动网络技术有限公司 用于生成信息的方法和装置
CN109558542B (zh) * 2018-12-04 2020-11-27 北京达佳互联信息技术有限公司 一种信息质量评估方法、信息推送方法及装置
CN109819015B (zh) * 2018-12-14 2022-08-19 深圳壹账通智能科技有限公司 基于用户画像的信息推送方法、装置、设备及存储介质
CN109862188B (zh) * 2019-02-12 2021-01-01 北京百度网讯科技有限公司 信息发送方法及装置、设备及存储介质
CN111475721B (zh) * 2020-03-31 2023-12-29 百度在线网络技术（北京）有限公司 信息推送方法、装置、设备及存储介质
EP3916585A4 (en) 2020-03-31 2022-04-13 Baidu Online Network Technology (Beijing) Co., Ltd INFORMATION OUTPUT METHOD AND DEVICE, DEVICE AND STORAGE MEDIUM
CN114422584B (zh) * 2021-12-16 2023-07-07 北京百度网讯科技有限公司 资源的推送方法、设备和存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104462593A (zh) * 2014-12-29 2015-03-25 北京奇虎科技有限公司 一种提供用户个性化资源消息推送的方法和装置
CN105159910A (zh) * 2015-07-03 2015-12-16 安一恒通（北京）科技有限公司 信息推荐方法和装置
CN105915438A (zh) * 2016-04-15 2016-08-31 北京奇虎科技有限公司 消息推送方法、装置及系统
CN106033419A (zh) * 2015-03-10 2016-10-19 深圳市腾讯计算机系统有限公司 一种消息实时推送方法、装置及系统
CN107222566A (zh) * 2017-07-06 2017-09-29 广州优视网络科技有限公司 信息推送方法、装置及服务器

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103679494B (zh) * 2012-09-17 2018-04-03 阿里巴巴集团控股有限公司 商品信息推荐方法及装置
CN103235823A (zh) * 2013-05-06 2013-08-07 上海河广信息科技有限公司 根据相关网页和当前行为确定用户当前兴趣的方法和系统
JP6235386B2 (ja) * 2014-03-19 2017-11-22 株式会社東芝 情報提示装置、情報提示方法およびプログラム
CN105183727A (zh) * 2014-05-29 2015-12-23 上海研深信息科技有限公司 一种图书推荐方法及其系统
CN104182516B (zh) * 2014-08-21 2018-03-06 北京金山安全软件有限公司 信息推荐方法、装置和移动终端
CN104182515B (zh) * 2014-08-21 2018-10-12 北京金山安全软件有限公司 信息推荐方法、装置和移动终端
CN105701182A (zh) * 2016-01-07 2016-06-22 百度在线网络技术（北京）有限公司 信息推送方法和装置
CN105872629B (zh) * 2016-03-18 2019-08-27 合一网络技术(北京)有限公司 内容推荐方法、装置及系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104462593A (zh) * 2014-12-29 2015-03-25 北京奇虎科技有限公司 一种提供用户个性化资源消息推送的方法和装置
CN106033419A (zh) * 2015-03-10 2016-10-19 深圳市腾讯计算机系统有限公司 一种消息实时推送方法、装置及系统
CN105159910A (zh) * 2015-07-03 2015-12-16 安一恒通（北京）科技有限公司 信息推荐方法和装置
CN105915438A (zh) * 2016-04-15 2016-08-31 北京奇虎科技有限公司 消息推送方法、装置及系统
CN107222566A (zh) * 2017-07-06 2017-09-29 广州优视网络科技有限公司 信息推送方法、装置及服务器

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110413895A (zh) * 2019-07-31 2019-11-05 腾讯科技（深圳）有限公司 一种信息推送的方法以及相关装置
CN111163152A (zh) * 2019-12-26 2020-05-15 中国平安人寿保险股份有限公司 信息推送方法及装置
CN111163152B (zh) * 2019-12-26 2023-08-08 中国平安人寿保险股份有限公司 信息推送方法及装置
CN112287236A (zh) * 2020-11-19 2021-01-29 每日互动股份有限公司 文本消息的推送方法、装置、计算机设备及存储介质
CN113312554A (zh) * 2021-06-15 2021-08-27 北京百度网讯科技有限公司 用于评价推荐系统的方法及装置、电子设备和介质
CN113312554B (zh) * 2021-06-15 2023-11-03 北京百度网讯科技有限公司 用于评价推荐系统的方法及装置、电子设备和介质

Also Published As

Publication number Publication date
CN108259546A (zh) 2018-07-06

Similar Documents

Publication Publication Date Title
WO2018130220A1 (zh) 消息推送方法、设备及可编程设备
US10310834B2 (en) Searching and accessing application functionality
TWI491217B (zh) 使用自使用者活動更新所得觀點增加內容關聯性之方法
US9256826B2 (en) Predicting reactions to short-text posts
US9626443B2 (en) Searching and accessing application functionality
JP6334696B2 (ja) ハッシュタグおよびコンテンツ提示
US10296641B2 (en) Techniques for efficient access of software application functionality in search
US20140095611A1 (en) Personalization through dynamic social channels
WO2018153179A1 (zh) 消息推送方法、设备及可编程设备
US10089652B2 (en) Generating advertisements for search results that reference software applications
US9953061B2 (en) Similarity engine for facilitating re-creation of an application collection of a source computing device on a destination computing device
US10289637B2 (en) Entity generation using queries
WO2018171271A1 (zh) 应用推荐方法、客户端、服务器及系统
CN103959745A (zh) 监视应用程序资源消耗
US20140123001A1 (en) User displays using n-way paginated merge of information from diverse sources
WO2018145609A1 (zh) 广告展示控制方法、设备、服务器及客户端
CN109978584A (zh) 一种信息流广告推送方法及装置
US20160012130A1 (en) Aiding composition of themed articles about popular and novel topics and offering users a navigable experience of associated content
US9471669B2 (en) Presenting previously selected search results
US10509831B2 (en) Systems and methods for time and space algorithm usage
US20160188721A1 (en) Accessing Multi-State Search Results
CA2864127A1 (en) Systems and methods for recommending advertisement placement based on in network and cross network online activity analysis
WO2018177415A1 (zh) 搜索结果推荐方法、设备、搜索引擎及电子设备和介质
WO2018205642A1 (zh) 视频收益计算建模装置、方法和视频推荐装置、方法和服务器及存储介质
Ziegler Radio as numbers: counting listeners in a big data world

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: 18739202

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 18/10/2019)

122 Ep: pct application non-entry in european phase

Ref document number: 18739202

Country of ref document: EP

Kind code of ref document: A1