CN114697688A - Live broadcast strategy recommendation method, interaction method, device, equipment and storage medium - Google Patents

Live broadcast strategy recommendation method, interaction method, device, equipment and storage medium Download PDF

Info

Publication number
CN114697688A
CN114697688A CN202011605791.2A CN202011605791A CN114697688A CN 114697688 A CN114697688 A CN 114697688A CN 202011605791 A CN202011605791 A CN 202011605791A CN 114697688 A CN114697688 A CN 114697688A
Authority
CN
China
Prior art keywords
live
live broadcast
process node
state index
action
Prior art date
Legal status (The legal status 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 status listed.)
Pending
Application number
CN202011605791.2A
Other languages
Chinese (zh)
Inventor
曹雅婷
孙大为
孙艳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Alibaba Group Holding Ltd
Original Assignee
Alibaba Group Holding Ltd
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 Alibaba Group Holding Ltd filed Critical Alibaba Group Holding Ltd
Priority to CN202011605791.2A priority Critical patent/CN114697688A/en
Publication of CN114697688A publication Critical patent/CN114697688A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/21—Server components or server architectures
    • H04N21/218—Source of audio or video content, e.g. local disk arrays
    • H04N21/2187—Live feed
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
    • H04N21/24—Monitoring of processes or resources, e.g. monitoring of server load, available bandwidth, upstream requests
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
    • H04N21/24—Monitoring of processes or resources, e.g. monitoring of server load, available bandwidth, upstream requests
    • H04N21/2407—Monitoring of transmitted content, e.g. distribution time, number of downloads
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
    • H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
    • H04N21/251—Learning process for intelligent management, e.g. learning user preferences for recommending movies
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
    • H04N21/442—Monitoring of processes or resources, e.g. detecting the failure of a recording device, monitoring the downstream bandwidth, the number of times a movie has been viewed, the storage space available from the internal hard disk
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
    • H04N21/442—Monitoring of processes or resources, e.g. detecting the failure of a recording device, monitoring the downstream bandwidth, the number of times a movie has been viewed, the storage space available from the internal hard disk
    • H04N21/44204—Monitoring of content usage, e.g. the number of times a movie has been viewed, copied or the amount which has been watched
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
    • H04N21/442—Monitoring of processes or resources, e.g. detecting the failure of a recording device, monitoring the downstream bandwidth, the number of times a movie has been viewed, the storage space available from the internal hard disk
    • H04N21/44213—Monitoring of end-user related data
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/45—Management operations performed by the client for facilitating the reception of or the interaction with the content or administrating data related to the end-user or to the client device itself, e.g. learning user preferences for recommending movies, resolving scheduling conflicts
    • H04N21/466—Learning process for intelligent management, e.g. learning user preferences for recommending movies
    • H04N21/4662—Learning process for intelligent management, e.g. learning user preferences for recommending movies characterized by learning algorithms
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/45—Management operations performed by the client for facilitating the reception of or the interaction with the content or administrating data related to the end-user or to the client device itself, e.g. learning user preferences for recommending movies, resolving scheduling conflicts
    • H04N21/466—Learning process for intelligent management, e.g. learning user preferences for recommending movies
    • H04N21/4667—Processing of monitored end-user data, e.g. trend analysis based on the log file of viewer selections
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
    • H04N21/45—Management operations performed by the client for facilitating the reception of or the interaction with the content or administrating data related to the end-user or to the client device itself, e.g. learning user preferences for recommending movies, resolving scheduling conflicts
    • H04N21/466—Learning process for intelligent management, e.g. learning user preferences for recommending movies
    • H04N21/4668—Learning process for intelligent management, e.g. learning user preferences for recommending movies for recommending content, e.g. movies

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Computing Systems (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Social Psychology (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The embodiment of the invention provides a live broadcast strategy recommendation method, an interaction method, a live broadcast strategy recommendation device, live broadcast strategy interaction equipment and a storage medium. The live broadcast strategy recommendation method comprises the following steps: acquiring a state index of a live broadcast process node; and matching the state index of the process node with a preset live broadcast state index, and determining a live broadcast strategy of the process node. The scheme of the embodiment of the invention realizes the real-time optimization of the live broadcast process and better assists the growth of the anchor broadcast.

Description

Live broadcast strategy recommendation method, interaction method, device, equipment and storage medium
Technical Field
The embodiment of the invention relates to the technical field of computers, in particular to a live broadcast strategy recommendation method, an interaction method, a live broadcast strategy recommendation device, a live broadcast strategy recommendation equipment and a storage medium.
Background
With the rapid development of network live broadcast, the number of anchor is also increased exponentially, and the continuous and stable growth of the anchor is an important consideration factor for various live broadcast platforms.
Generally, the most important growth aspect of the anchor is from the control of the live broadcast, which requires the anchor student to adjust his own dialect, interaction strategy, etc. in time according to the online condition of the fans in the live broadcast and the comprehensive factors of each index of the interaction condition, so as to gradually grow into an excellent anchor.
However, in the prior art, the anchor generally learns according to own experience, and the anchor has a slow growth rate.
Disclosure of Invention
In view of this, embodiments of the present invention provide a live broadcast policy recommendation method, an interaction method, an apparatus, a device and a storage medium, so as to solve or alleviate the above problems.
According to a first aspect of an embodiment of the present invention, a live broadcast policy recommendation method is provided, including: acquiring a state index of a live broadcast process node; and matching the state index of the process node with a preset live broadcast state index, and determining a live broadcast strategy of the process node.
According to a second aspect of the embodiments of the present invention, there is provided a policy recommendation method, including: acquiring a state index of a recorded broadcast process node; and matching the state index of the process node with a preset live broadcast state index, and determining the interaction strategy of the process node.
According to a third aspect of the embodiments of the present invention, there is provided an interaction method, including: determining a status index of a live process node, wherein the status index of the process node is used for being matched with a preset live broadcast status index so as to determine a live broadcast strategy of the process node; and acquiring a live broadcast strategy of the process node.
According to a fourth aspect of the embodiments of the present invention, there is provided a live broadcast policy recommendation apparatus, including: the state index acquisition module is used for acquiring the state index of the live broadcast process node; and the live broadcast strategy determining module is used for matching the state index of the process node with a preset live broadcast state index and determining the live broadcast strategy of the process node.
According to a fifth aspect of the embodiments of the present invention, there is provided a policy recommendation apparatus including: the state index acquisition module is used for acquiring the state index of the recorded and broadcast process node; and the strategy determining module is used for matching the state index of the process node with a preset live broadcast state index and determining the interaction strategy of the process node.
According to a sixth aspect of the embodiments of the present invention, there is provided an interactive apparatus, including: the system comprises a status index determining module, a live broadcast strategy determining module and a live broadcast strategy determining module, wherein the status index of a live broadcast process node is determined and is used for being matched with a preset live broadcast status index so as to determine a live broadcast strategy of the process node; and the live broadcast strategy acquisition module is used for acquiring the live broadcast strategy of the process node.
According to a seventh aspect of embodiments of the present invention, there is provided an electronic apparatus, the apparatus including: the system comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete mutual communication through the communication bus; the memory is used for storing at least one executable instruction, and the executable instruction causes the processor to execute the corresponding operation of the method according to the first aspect or the second aspect.
According to an eighth aspect of embodiments of the present invention, there is provided a storage medium having stored thereon a computer program which, when executed by a processor, implements a method as claimed in the first or second aspect.
In the scheme of the embodiment of the invention, the state index of the process node can be matched with the preset live broadcast state index, and the live broadcast strategy of the process node is determined, so that the real-time optimization of the live broadcast process is realized, and the growth of the anchor is better assisted.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the embodiments of the present invention, and it is also possible for a person skilled in the art to obtain other drawings based on the drawings.
FIG. 1 is a schematic diagram of an exemplary network architecture suitable for live video;
FIG. 2 is a schematic flow chart diagram of a live policy recommendation method according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a live policy recommendation method according to another embodiment of the present invention;
FIG. 4 is a schematic flow chart diagram of an interaction method according to another embodiment of the present invention;
fig. 5 is a schematic block diagram of a live policy recommendation apparatus according to another embodiment of the present invention;
FIG. 6 is a schematic block diagram of an interactive apparatus according to another embodiment of the present invention;
FIG. 7 is a schematic flow chart diagram of a policy recommendation method of one embodiment of the present invention;
FIG. 8 is a schematic block diagram of a policy recommendation device according to another embodiment of the present invention;
fig. 9 is a hardware configuration of an electronic device according to another embodiment of the present invention.
Detailed Description
In order to make those skilled in the art better understand the technical solutions in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all the embodiments. All other embodiments obtained by a person skilled in the art based on the embodiments in the embodiments of the present invention should fall within the scope of protection of the embodiments of the present invention.
The following further describes specific implementation of the embodiments of the present invention with reference to the drawings. Fig. 1 is a schematic diagram of an exemplary network architecture suitable for live video. User device 100 includes a network interface 110, a live video application 120, and an interpersonal interaction interface 130. The live server 200 comprises a network interface 210 and live data, e.g. for background analysis of live data or live data presented at the user device 100 side. The live video application 120 may be a browser application, a video-type playing application, or other application with live video functionality, etc. The live video application 120 may be installed in an operating system of the user equipment 100, or may be installed in other applications in a manner such as an SDK (Software Development Kit).
In general, the anchor 300 may experience its own experience to gain technical growth based on live data presented to the user device 100 by the live server 200. However, this approach makes the growth rate of the anchor slower.
Fig. 2 is a schematic flowchart of a live policy recommendation method according to an embodiment of the present invention. The solution of the present embodiment may be applied to any suitable electronic device with data processing capability, including but not limited to: live-enabled servers such as live servers or private, public, or hybrid clouds. The live broadcast policy recommendation method of fig. 2 includes:
210: and acquiring the state index of the live process node.
It should be understood that the status indicators of the live process nodes include, but are not limited to, the number of people online, the length of time online, the number of praise, the number of bean vermicelli remaining, the number of newly added bean vermicelli, the number of times shopping cart addition is guided, the number of collection people is guided, and the amount of bargaining is guided. In addition, the duration of the live broadcast period is not limited. In the live broadcast period, the process nodes can be interactive nodes or non-interactive nodes. The anchor can perform multiple live interactive actions at the interactive node. The number of interactive actions of the anchor may be determined based on the duration of the live period, or the number of interactive actions of the live may be associated with the live duration.
220: and matching the state index of the process node with a preset live broadcast state index, and determining a live broadcast strategy of the process node.
It should be appreciated that the live strategy of a process node can be determined through reinforcement learning using the large amount of data accumulated by the live platform. Specifically, the live broadcast state index and the corresponding interactive action point in the live broadcast process of the deep reinforcement learning real-time analysis can be matched with the state index of the process node and the preset live broadcast state index, so that the anchor is assisted to recommend the interactive strategy which should be adopted by the live broadcast at present, and whether the anchor adopts the recommendation strategy or not is monitored in real time.
It should also be appreciated that the reinforcement learning described above may be implemented using a reinforcement learning algorithm, such as a Q-value learning algorithm, for example, a Deep Q-learning Network (DQN) algorithm. For example, the current learning target value is the current Q value. For example, for multiple interactive actions, previous each interactive action and previous each live-room state may be determined, and the current learning objective value is updated based on the previous each interactive action and previous each live-room state and the state index contribution value of the current interactive action.
In other examples, other reinforcement learning or supervised learning algorithms may also be employed to implement embodiments of the present invention.
In the scheme of the embodiment of the invention, the state index of the process node can be matched with the preset live broadcast state index, and the live broadcast strategy of the process node is determined, so that the real-time optimization of the live broadcast process is realized, and the growth of the anchor is better assisted. In other words, the recommendation algorithm is optimized and improved due to the monitored status indicators. The anchor can optimize own talk and interaction strategies in real time in the live broadcast process, and continuously accumulate experiences, thereby achieving rapid growth.
In one example, a status indicator of a process node of a live of a target live room parameter may be obtained. Target live broadcast room parameters include, but are not limited to, live broadcast room types such as game live broadcast room, live tape live broadcast room, teaching live broadcast room, and the like. The parameters of the live broadcast room further comprise live broadcast room list such as a main broadcast list and a visitor list.
In another example, a status indicator of a live process node may be obtained, the status indicator including at least one of a number of sales of a particular item, a number of praise's, a number of directed shopping cart additions, an amount of deals, or an amount of deals. The state indexes further comprise a merchant ranking list, a commodity ranking list, a brand ranking list and the like. The status indicators of the process nodes may also include a fan image and a cast image. The anchor portrait comprises the expression type of the anchor and the expression of the interactive fan.
In a live educational scenario, a (e.g., server) can obtain an instructional state indicator for a process node of an instructional live. And matching the state index of the process node with a preset teaching state index to determine a teaching interaction strategy of the process node. It should be understood that the teaching status indicators include the attendance rate of students, the frequency of questions asked by students, the liveness of students in class, the accuracy of student answers, and the like.
In one example, the server may also match the status indicator of the current process node with a preset teaching status indicator, determine a first teaching interactive operation strategy for the teacher account and a second teaching interactive operation strategy for the student account. The server can recommend corresponding teaching interaction operations to the teacher account (such as the first terminal device where the teacher account is located) and the student account (such as the second terminal device where the student account is located). The server can record corresponding teacher historical interaction operation parameters and student historical interaction operation parameters aiming at the teacher account or the student account.
In a social live broadcast scenario, a social status index of a process node of the social live broadcast may be obtained. And matching the state index of the process node with a preset social state index to determine the social interaction strategy of the process node. It should be understood that social live includes live with goods, live talent, live chat, and the like.
In one example, the server may also match the status indicator of the current process node with a preset social status indicator, determine a first interactive operation policy for the anchor account and a second interactive operation policy for the fan account. The server may recommend a corresponding social interaction operation to the anchor account (e.g., the first terminal device in which it is located) and the fan account (or viewer account) (e.g., the second terminal device in which it is located). The server can record corresponding teacher historical interaction operation parameters and student historical interaction operation parameters aiming at the anchor account or the fan account. The server can also select to record or recommend the interactive operation strategy for the members higher than the specific level, and not record or recommend the interactive operation strategy for the members lower than the specific level. The server may award fan account numbers with higher interaction indices (indices indicating frequency of interaction).
In another example, the emulated interactive operation may be recommended to the anchor account (by sending a message to the first terminal device) by a status indicator of a process node in the live broadcast room, and after the anchor performs a specific action, the emulated interactive operation corresponding to the emulated interactive operation may be recommended to the fan account (by sending a message to the second terminal device). The server may maintain data corresponding to status indicators of each process node of the live broadcast room, or the server may maintain data corresponding to status indicators of each process node of the anchor account or the fan account. The server can maintain the corresponding relation between the dynamic anchor recommended interactive operation strategy and the recommended interactive operation strategy of the specific fan. The server can determine the corresponding relation according to the interaction index (including interaction frequency, reward amount, transaction amount and the like) of the anchor and the specific fan, so that the interaction index is positively correlated with the association degree between the anchor recommended interaction operation strategy and the recommended interaction operation strategy of the specific fan, wherein the association degree indicates the ratio of the recommended interaction operation duration of the specific fan to the anchor recommended interaction operation duration, for example, for anchor simulation interaction operation recommended to the fan account number, if the participation duration of the specific fan is long, the association degree with the anchor account number is large. If the participation time of the specific fan is short, the association degree with the anchor account is small.
In another example, a video access operation may be recommended to the anchor account (by sending a message to the first terminal device) according to a status indicator of a process node in the live broadcast room, and after the anchor performs a complete video access introduction operation, an interactive operation of accessing a video may be recommended to a specific fan account (by sending a message to the second terminal device), for example, the video of the specific fan is accessed to a live broadcast main screen so as to be seen by other fans. In addition, questions can be recommended to the anchor account and the specific fan account, or the server can be accessed to the intelligent question and answer server, and the server can access the question and answer database. For example, a corresponding next question list or answer list may be recommended to the primary account (by sending a message to the first terminal device) in accordance with the answer or question for a particular fan account number. For example, a corresponding next question list or answer list may also be recommended to a particular fan account number (by sending a message to the second terminal device) based on the answer or question of the anchor account number.
In another implementation of the invention, the method further comprises: and recommending live action candidates according to the live strategy of the process node.
In another implementation of the invention, the method further comprises: determining a selected action in the live action candidates; judging the state index contribution value of the selected action to the process node; and updating the live action candidates according to the contribution values.
Specifically, in the reinforcement learning described above, the current learning target value may be determined based on the previous interactive action in the live broadcast cycle and the previous live broadcast room state. A current recommended action for the current interactive action may be recommended based on the current learning objective value. The current learning target value may be updated based at least on the state index contribution value of the current interactive action. The current learning target value can be determined according to the previous interactive action and the previous live broadcasting room state in the live broadcasting period, so that the overall optimization of the live broadcasting process is realized, and the growth of the anchor is better assisted.
It should also be appreciated that the decision algorithm for the point in time of each of the plurality of interactive actions may be obtained based on a large amount of historical data. In one example, the time point of each interactive action may also be determined based on at least one of a duration of the target session being live, a number of times the interactive action was live a plurality of times, and a status indicator contribution value of a last interactive action.
It is also understood that the previous interactive action may be at least one interactive action prior to the current interactive action. The previous live room state may be a live room state formed after the previous interaction. The live room state can be characterized in at least one dimension.
It should be appreciated that recommending the current recommendation action may include sending recommendation data to a user device of the front end. The user equipment can generate a recommendation control based on the recommendation data and a frame of a live interface of the user equipment. The interaction between the live broadcast server and the user equipment may be based on the network architecture shown in fig. 1, or may be based on other network architectures.
It should be understood that the current interactive action may be selected from the recommended actions, or may be an interactive action that is executed by the anchor and is unrelated to the recommended action, which is not limited in the embodiment of the present invention.
As an example, a plurality of learning target values may be determined according to a plurality of historical interactive actions in a live broadcast cycle and a plurality of live broadcast room states corresponding to the plurality of historical interactive actions; and performing weighting processing on the plurality of learning targets to obtain the current learning target value. Because the current learning target value is obtained by weighting the plurality of learning targets, the data processing efficiency is improved in the process of real-time recommendation.
In another implementation of the present invention, determining the state index contribution value of the selected action to the process node includes: determining a group of state index dimensions, wherein the group of state index dimensions comprise at least one of average online time, newly-added vermicelli number, store introduction and total transaction amount; and judging the contribution value of the selected action to the state index of the process node under a set of state index dimensions.
The adoption of the state index contribution value based on the indexes of the various state index contribution values is beneficial to improving the state index contribution values of multiple dimensions, so that the efficiency of reinforcement learning is improved, and the speed of the growth of the anchor is further improved.
As one example, a component status indicator contribution value for a component status indicator contribution value indicator may be determined for the current interaction action; and (4) performing superposition processing on the component state index contribution values to obtain a total state index contribution value so as to update the current learning target value. The total state index contribution value is obtained by overlapping the component state index contribution values to update the current learning target value, and different component state index contribution values can reflect the conditions of different component state index contribution value indexes, so that a more accurate total state index contribution value is obtained in a live broadcast scene.
As an example, a set of live action candidates for a current interactive action may be recommended according to a current learning objective value, wherein the current interactive action includes at least one interactive action of the set of live action candidates. Further, a multi-component status indicator contribution value may be determined for each of the at least one interactive action corresponding to the plurality of component status indicator contribution value indicators; and respectively summing the multi-component state index contribution values to obtain a plurality of component state index contribution values.
As one example, it may be determined that the current interaction action corresponds to an initial state index contribution value and a cost value for each of a set of state index contribution value indices; and determining the sub-state index contribution value of the current interaction action corresponding to the sub-state index contribution value index by removing the cost value from the initial sub-state index contribution value corresponding to each sub-state index contribution value index to obtain a group of sub-state index contribution values. Because the cost value is removed from the initial state index contribution value, the obtained state index contribution value is more accurate, and thus a more accurate total state index contribution value is obtained.
As one example, a set of live action candidates for a current interactive action may be recommended based on a current learning objective value. Because a group of live action candidates for the current interactive action is recommended, more flexible selection of the interactive action is improved for the anchor, the interactive effect is improved, and the growth of the anchor is facilitated.
In another implementation manner of the present invention, determining a selected action in the live action candidates includes: and determining the selected action in the live action candidates by performing multi-modal recognition on the target operation. For example, by performing multi-modal recognition on the target operation, the obtained recognition result is determined as the current interactive action. Because the current interactive action which does not belong to a group of live action candidates is obtained in a multi-mode recognition mode, the flexibility of the recommendation method is improved.
In another implementation of the present invention, determining a selected action in a set of live action candidates comprises: the selected action is determined by monitoring each live action candidate in a set of live action candidates. For example, the at least one interactive action is acquired by monitoring a trigger operation of the at least one interactive action. By monitoring the manner in which the trigger operation is triggered, at least one interactive action is reliably captured.
In another implementation manner of the present invention, determining a selected action in the live action candidates includes: determining a selected action in a group of live action candidate sets, wherein the group of live action candidate sets comprises at least one live action candidate of red pack, prize drawing, on-air commodity shelving, special coupons, attention releasing small cards, shop card releasing, mouth-air guide attention and mouth-air question-answer interaction. Because the interaction mode can be selected from various live action candidates for recommendation, the state index contribution values of multiple dimensions can be improved, and the growth speed of the anchor is further improved. It should be appreciated that individual dimension values in a set of live room state dimensions subsequent to a previous interaction action may also be determined and the individual dimension values determined as previous live room states to improve the accuracy of the live room states.
In another implementation of the invention, the method further comprises: and performing optimization adjustment on a group of live action candidate sets.
In another implementation of the invention, the method further comprises: establishing a live broadcast state index pool, wherein the live broadcast state index pool comprises a plurality of live broadcast state indexes; the live broadcast status indexes are as follows: the number of online people, the online time, the number of praise, the number of vermicelli reserved, the number of newly added vermicelli, the number of times of guiding to add a shopping cart, the number of collecting people and the amount of guided bargaining. Due to the fact that the multiple state indexes are adopted, the more accurate growth live broadcast state can be obtained, the efficiency of reinforcement learning is improved, and the speed of anchor growth is further improved. Specifically, when the anchor confirms the target growth live broadcast state, one or more indexes of the target growth live broadcast state can be selected to generate a multi-dimensional state matrix for processing.
In another implementation manner of the present invention, the preset live broadcast status index may be determined as follows: determining a preset live broadcast state index according to the selection of a user in the live broadcast state index pool; the preset live broadcast state indexes are one or more.
As an example, the current learning target value of the current interactive action may be determined according to the state index contribution value of the last interactive action. The current interactive action can be recommended according to the current learning target value, and the learning target value of the next interactive action is determined based on the state index contribution value of the current interactive action. Because the current learning target value of the current interactive action is determined based on the state index contribution value of the previous interactive action, and the learning target value of the next interactive action is determined based on the state index contribution value of the current interactive action, each interactive action is related to the previous state index contribution value, thereby improving the effect of dynamic reinforcement learning.
In other words, real-time decision recommendation is performed in a live broadcast scene by means of an algorithm, live broadcast strategy recommendation is not confirmed by a single time node, the whole live broadcast overall effect is taken as a final target, the interaction mode of each node is comprehensively recommended, and the overall effect maximization is achieved.
As an example, an interaction time node and a set of live action candidates of the current interaction action may be determined according to the current learning target value; a set of live action candidates is recommended before an interaction time node for a current interaction action. Because the interaction time node and the live action candidate are beneficial to improving the state index contribution values of multiple dimensions, the growth speed of the anchor is further improved.
As an example, a set of live action candidates may be determined according to an interaction time node of a current interaction action. Because a group of live action candidates are determined according to the interaction time node of the current interaction action, the time node is related to the live action candidates, and therefore the efficiency of reinforcement learning is improved.
As one example, "red envelope," "drawing prizes," "special coupons," etc. are interactive actions with an operating cost to the anchor. The recommendation priority may be determined based on the interactive actions, or model optimization may be performed using costs of the interactive actions as constraints. For example, the recommendation weight coefficient for the interactive action may be adjusted lower based on the cost of the interactive action.
In other words, a set of live action candidates presents an interactive approach in a tasking rather than a single tasking, while conditionally constraining the scores of the interactive actions with costs and rewards to find a policy optimization solution. Due to the adoption of the mode of task-division superposition condition constraint, the effect of the scheme of the invention can be more suitable for the actual situation and complexity of a live broadcast scene.
Further, in one example, the tasking state indicator contribution value may generate a first weight value according to an indicator selected by a user, and the unselected indicator defaults to a second weight value, wherein the first weight value is greater than the second weight value.
As one example, a set of response results for live action candidates may be obtained; based on the response result, a state index contribution value of the current interactive action is determined.
As an example, an interaction mode selected from a set of live action candidates may be obtained; or, collecting the interaction modes which do not belong to a group of live action candidates.
Due to the fact that the interaction mode selected from the group of live action candidates is obtained, subsequent reinforcement learning and recommendation can be conducted based on the effect of the selected interaction mode, and therefore accuracy of dynamic reinforcement learning is improved.
Because the collection does not belong to a group of live action candidate interaction modes, the effects of other interaction modes can be collected, so that the learning samples are richer, and the accuracy of dynamic reinforcement learning is improved.
As an example, a buried site may be required to monitor in real time whether the anchor has taken a corresponding decision. In another example, it may be monitored through multimodal recognition whether the anchor has taken a corresponding decision. Based on which subsequent state index contribution value calculations and target value updates are performed.
Since the default agent in this example will not necessarily implement decision suggestions, the recommended and monitored solution is closer to the actual scenario.
As one example, current variation information for a component status indicator contribution value indicator may be determined based on the response result; and determining the contribution value of the state index of the current interaction action based on the current change information. As one example, a last-grown-live status of a last interactive action at a anchor growth target may be determined; determining the current growth live broadcast state of the current interactive action at the anchor growth target based on the last growth live broadcast state and the state index contribution value of the last interactive action; and determining a current learning target value of the current interaction action relative to the current growth live state. As one example, a set of live action candidates may also be determined based on a current growing live status.
As one example, the last grown live status and the current grown live status are based on a set of live room status dimensions, wherein the last anchor grown live status comprises last grown live status values for status indicators in the set of live room status dimensions and the current anchor grown live status comprises current grown live status values for status indicators in the set of live room status dimensions.
Because the live state of last growth and the live state of current growth are based on a set of live room state dimension, consequently, improved more accurate live state of growth.
Fig. 3 is a schematic diagram of a live policy recommendation method according to another embodiment of the present invention. As shown, in step S311, the core growth target of the anchor is determined, and the process proceeds to step S312. In particular, for a host, the core growth goals may themselves focus on the core goals of interest. The core targets may be the same when the anchor is in different lifecycle stages.
Further, in one example, the growth goal may be that the anchor selects by its own preference. In another example, the other is to determine the default core growth target for different levels of the anchor in different industries by analyzing the historical growth path of the anchor of the platform through the big data of the platform. In the embodiment of the invention, at least one of the online time length, the number of newly added fans and the guided transaction amount can be used as an anchor growth target. For example, the core of the newcomer anchor is the number of newly-increased fans, the core of the anchor below the middle waist is the online time length, and the core of the anchor above the middle waist is the guided transaction amount.
In step S312, the agent assists in popping up the interactive decision suggestion, and proceeds to step S313.
In step S313, it is detected whether the anchor takes a policy, and it proceeds to step S314. Specifically, after the anchor completes the selection, the growth target that needs to be defined for real-time online decision making is determined. And calling the decision recommendation learned by the deep learning algorithm based on the given growth target in real time in the live broadcast, and giving the live broadcast strategy recommendation in the live broadcast, wherein the specific live broadcast strategy recommendation can be formed by a DQN algorithm based on a feasible interactive action point decision in all interactive action candidate sets. Meanwhile, a live broadcast usually lasts for several hours.
In addition, different time nodes and different growth target states of the live broadcast rooms have different recommended interaction actions, and in the embodiment of the invention, the decision recommendation of each node is carried out by the whole index optimization of the complete live broadcast field instead of carrying out decision suggestion by single-point optimization of each single node.
In step S314, the reward for the current interactive action is calculated, and the process proceeds to step S315.
In step S315, the interaction target value is updated, and the process proceeds to step S316. Specifically, after a decision suggestion is given by a live broadcasting agent, the system needs to record and calculate two indexes, one is whether a main broadcasting takes a decision interaction action, and the other is to calculate a state index contribution value and a target Q value after a decision is taken in real time to update. The record of whether the anchor really adopts the recommended interaction action can be confirmed by clicking the embedded point through the action point, and for the fact that the interface broadcast type can not be obtained from the clicked data, the record can be identified and confirmed in a multi-mode identification mode. The calculation of the state index contribution value and the calculation of the Q value are carried out by a DQN algorithm formula, wherein the scheme adopts multi-objective optimization, and the reason is that when decision suggestion is carried out on a complete live broadcast period, for each node, the single-objective state index contribution value is not enough to completely describe the change of a live broadcast room, and the multi-objective optimization is adopted, and the real-time decision effect of the complete session can be better ensured by setting different weights (w) of each objective according to different growth objectives.
In step S316, the core growth goal of the anchor is determined, and the process returns to step S312 to resume the recommendation of the next interaction strategy. Furthermore, in one example, after the live broadcast is over, the decision execution rate detected in real time and the performance statistics of the overall live broadcast can be presented to the anchor.
Further, in another example, the boosting of the algorithm may be based on an analysis of the effect of the decision execution.
It is to be understood that the processes of step S312 to step S315 may be realized by an algorithm such as a Q-value learning algorithm.
Fig. 4 is a schematic block diagram of an interaction method according to another embodiment of the invention. The solution of the present embodiment may be applied to any suitable electronic device with data processing capability, including but not limited to: a terminal device such as a tablet computer, a desktop computer or any other device having interactive functionality such as live video. The interaction method of fig. 4 includes:
410: and determining the status index of the live process node, wherein the status index of the process node is used for being matched with a preset live broadcast status index so as to determine the live broadcast strategy of the process node.
420: and acquiring a live broadcast strategy of the process node.
The adoption of the state index contribution value based on the indexes of the various state index contribution values is beneficial to improving the state index contribution values of multiple dimensions, so that the efficiency of reinforcement learning is improved, and the speed of the growth of the anchor is further improved.
In addition, the embodiment of the invention can adopt a large amount of data accumulated by a live broadcast platform, assist the anchor in recommending the interaction strategy which should be adopted at present by deeply and intensively learning various data indexes and corresponding interaction action points in the live broadcast process of real-time analysis, and monitor whether the anchor adopts the recommendation strategy or not in real time.
In one example, the terminal device obtains a live policy of the process node, including: a terminal device receives a set of live action candidate sets. The set of live action candidates comprises at least one live action candidate of a red packet, a prize extraction, a goods on-air listing, a special coupon, an attention-paid small card, a shop card, an attention-directed mouth-broadcast, and a question-answer interaction.
In another example, the method further comprises: after receiving a set of live action candidate sets, a display control of the set of live action candidate sets is presented. The anchor may perform a selected live action of a set of live action candidates in response to a trigger operation of the display control. The live action of the display content performed by the anchor based on the display control can also be monitored.
The terminal equipment can send the parameters of the live action to a server with a live action function, and the server determines the learning target value of the current process node based on the state indexes corresponding to the parameters of the live action. And the server matches the state index of the current process node with a preset live broadcast state index to determine a live broadcast strategy of the next process node.
In one example, the server sends visualization information corresponding to parameters of the live action to the terminal device, wherein the visualization information includes dynamic rendering parameters of the live action. The terminal equipment comprises first terminal equipment of a main account and second terminal equipment of a spectator account. The server determines dynamic rendering parameters indicating different effects of the live action according to different member levels of the viewer account. In other words, the server determines a dynamic rendering parameter indicating a live action effect corresponding to the member level of the target audience account, and sends the dynamic rendering parameter to the second terminal device of the target audience account, so that users with different member levels can see different dynamic live action effects, and the participation degree of the users is improved.
Fig. 5 is a schematic block diagram of a live policy recommendation apparatus according to another embodiment of the present invention. The solution of the present embodiment may be applied to any suitable electronic device with data processing capability, including but not limited to: live-enabled servers such as live servers or private, public, or hybrid clouds. The live broadcast policy recommendation apparatus of fig. 5 includes:
the status indicator obtaining module 510 obtains a status indicator of a live process node.
The live broadcast policy determining module 520 matches the status index of the process node with a preset live broadcast status index to determine a live broadcast policy of the process node.
The adoption of the state index contribution value based on the indexes of the various state index contribution values is beneficial to improving the state index contribution values of multiple dimensions, so that the efficiency of reinforcement learning is improved, and the speed of the growth of the anchor is further improved.
In another implementation of the present invention, the apparatus further comprises: and the recommending module recommends live action candidates according to the live strategy of the process node.
In another implementation of the present invention, the apparatus further comprises: the determining module is used for determining the selected action in the live action candidates; the judgment module is used for judging the state index contribution value of the selected action to the process node; and the updating module updates the live action candidates according to the contribution values.
In another implementation manner of the present invention, the determining module is specifically configured to: determining a selected action in a group of live action candidate sets, wherein the group of live action candidate sets comprises at least one live action candidate of red pack, prize drawing, on-air commodity shelving, special coupons, attention releasing small cards, shop card releasing, mouth-air guide attention and mouth-air question-answer interaction.
In another implementation manner of the present invention, the determining module is specifically configured to: the selected action is determined by monitoring each live action candidate in a set of live action candidates.
In another implementation manner of the present invention, the determining module is specifically configured to: and determining the selected action in the live action candidates by performing multi-modal recognition on the target operation.
In another implementation manner of the present invention, the determining module is specifically configured to: determining a group of state index dimensions, wherein the group of state index dimensions comprise at least one of average online time, newly-added vermicelli number, store introduction and total transaction amount; and judging the contribution value of the selected action to the state index of the process node under a set of state index dimensions.
In another implementation of the present invention, the apparatus further comprises: the establishing module is used for establishing a live broadcast state index pool, the live broadcast state index pool comprises a plurality of live broadcast state indexes, and the live broadcast state indexes are as follows: the number of online people, the online time, the number of praise, the number of stored vermicelli, the number of newly added vermicelli, the number of times of guiding to add a shopping cart, the number of collection people and the amount of finished deals.
In another implementation manner of the present invention, the determining module is further configured to determine a preset live broadcast status indicator according to a selection of a user in the live broadcast status indicator pool; the preset live broadcast state indexes are one or more.
The apparatus of this embodiment is used to implement the corresponding method in the foregoing method embodiments, and has the beneficial effects of the corresponding method embodiments, which are not described herein again. In addition, the functional implementation of each module in the apparatus of this embodiment can refer to the description of the corresponding part in the foregoing method embodiment, and is not described herein again.
Fig. 6 is a schematic block diagram of an interaction device according to another embodiment of the invention. The solution of the present embodiment may be applied to any suitable electronic device with data processing capability, including but not limited to: a terminal device such as a tablet computer, a desktop computer or any other device having interactive functionality such as live video. The interactive apparatus of fig. 6 includes:
the status indicator determining module 610 determines a status indicator of a live process node, where the status indicator of the process node is used to match with a preset live status indicator to determine a live policy of the process node.
And a live broadcast strategy acquisition module 620 for acquiring the live broadcast strategy of the process node.
The adoption of the state index contribution value based on the indexes of the various state index contribution values is beneficial to improving the state index contribution values of multiple dimensions, so that the efficiency of reinforcement learning is improved, and the speed of the growth of the anchor is further improved.
The apparatus of this embodiment is used to implement the corresponding method in the foregoing multiple method embodiments, and has the beneficial effects of the corresponding method embodiment, which are not described herein again. In addition, the functional implementation of each module in the apparatus of this embodiment can refer to the description of the corresponding part in the foregoing method embodiment, and is not described herein again.
FIG. 7 is a schematic flow chart diagram of a policy recommendation method according to an embodiment of the present invention. The solution of the present embodiment may be applied to any suitable electronic device with data processing capability, including but not limited to: live-enabled servers such as live servers or private, public, or hybrid clouds. The policy recommendation method of fig. 7 includes:
710: acquiring a state index of a recorded broadcast process node;
720: and matching the state index of the process node with a preset live broadcast state index to determine the interaction strategy of the process node.
It should be understood that the interaction strategy of the process node is determined by matching the status index of the recorded and broadcast process node with the preset live broadcast status index, so that the interaction of the recorded and broadcast process is realized.
In one example, the server may record historical interactive operation parameters during a live broadcast process, where the historical interactive operation parameters include a main broadcast operation time parameter, a main broadcast interactive operation content parameter, an audience operation time parameter, an audience interactive operation content parameter, a correspondence between audience interactive operation content and main broadcast interactive operation content, and the like. For example, when the time indicated by the operation time parameter corresponding to the recorded and broadcast current process node is determined to arrive, the interactive operation content parameter is obtained, and the interactive strategy of the process node is determined according to the interactive operation content parameter. Further, the server may also send a control message indicating the interaction policy to the terminal device.
In another example, the anchor account may be logged in at a first terminal device and the viewer account may be logged in at a second terminal device. The server may obtain the interactive operation from the viewer account and the anchor account. The server can be used for viewing parameters of interactive operation of the account, and the parameters of the interactive operation can comprise an operation time parameter and an operation content parameter. The server can determine the anchor operation time parameter and the anchor interaction operation content parameter according to the corresponding relation between the audience interaction operation content and the anchor interaction operation content, and recommend an interaction strategy to the second terminal equipment at the time node indicated by the anchor operation time parameter.
Fig. 8 is a schematic block diagram of a policy recommendation apparatus according to another embodiment of the present invention. The solution of the present embodiment may be applied to any suitable electronic device with data processing capability, including but not limited to: live-enabled servers such as live servers or private, public, or hybrid clouds. The policy recommendation apparatus of fig. 8 includes:
the status index acquisition module 810 acquires status indexes of recorded and broadcast process nodes;
the policy determination module 820 matches the status indicator of the process node with a preset live broadcast status indicator to determine an interaction policy of the process node.
Fig. 9 is a hardware configuration of an electronic apparatus according to another embodiment of the present invention; as shown in fig. 9, the hardware structure of the electronic device may include: a processor 901, a communication interface 902, a storage medium 903 and a communication bus 904;
the processor 901, the communication interface 902 and the storage medium 903 are communicated with each other through a communication bus 904;
alternatively, the communication interface 902 may be an interface of a communication module;
the processor 901 may be specifically configured to: acquiring a state index of a live broadcast process node; matching the state index of the process node with a preset live broadcast state index, and determining a live broadcast strategy of the process node;
or determining a status index of a live process node, wherein the status index of the process node is used for being matched with a preset live broadcast status index so as to determine a live broadcast strategy of the process node; and acquiring a live broadcast strategy of the process node.
The Processor may be a general-purpose Processor, and includes a Central Processing Unit (CPU), a Network Processor (NP), and the like; but may also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components. The various methods, steps and logic blocks disclosed in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The storage medium may be, but is not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), a Programmable Read Only Memory (PROM), an Erasable Read Only Memory (EPROM), an electrically Erasable Read Only Memory (EEPROM), and the like.
In particular, the processes described above with reference to the flow diagrams may be implemented as computer software programs, according to embodiments of the present disclosure. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a storage medium, the computer program comprising program code configured to perform the method illustrated by the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network via the communication section, and/or installed from a removable medium. The computer program performs the above-described functions defined in the method of the present invention when executed by a Central Processing Unit (CPU). It should be noted that the storage medium of the present invention can be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access storage media (RAM), a read-only storage media (ROM), an erasable programmable read-only storage media (EPROM or flash memory), an optical fiber, a portable compact disc read-only storage media (CD-ROM), an optical storage media piece, a magnetic storage media piece, or any suitable combination of the foregoing. In the present invention, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In the present invention, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any storage medium that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a storage medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
Computer program code configured to carry out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may 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 and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may operate over any of a variety of networks: including a Local Area Network (LAN) or a Wide Area Network (WAN) -to the user's computer, or alternatively, to an external computer (e.g., through the internet using an internet service provider).
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions configured to implement the specified logical function(s). In the above embodiments, specific precedence relationships are provided, but these precedence relationships are only exemplary, and in particular implementations, the steps may be fewer, more, or the execution order may be modified. That is, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The modules described in the embodiments of the present invention may be implemented by software or hardware. The names of these modules do not in some cases constitute a limitation of the module itself.
As another aspect, the present invention also provides a storage medium having stored thereon a computer program which, when executed by a processor, implements the method as described in the above embodiments.
As another aspect, the present invention also provides a storage medium, which may be contained in the apparatus described in the above embodiments; or may be present separately and not assembled into the device. The storage medium carries one or more programs that, when executed by the apparatus, cause the apparatus to: acquiring a state index of a live broadcast process node; matching the state index of the process node with a preset live broadcast state index, and determining a live broadcast strategy of the process node;
or determining a status index of a live process node, wherein the status index of the process node is used for being matched with a preset live broadcast status index so as to determine a live broadcast strategy of the process node; and acquiring a live broadcast strategy of the process node.
The expressions "first", "second", "said first" or "said second" used in various embodiments of the present disclosure may modify various components regardless of order and/or importance, but these expressions do not limit the respective components. The above description is only configured for the purpose of distinguishing elements from other elements. For example, the first user equipment and the second user equipment represent different user equipment, although both are user equipment. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of the present disclosure.
When an element (e.g., a first element) is referred to as being "operably or communicatively coupled" or "connected" (operably or communicatively) to "another element (e.g., a second element) or" connected "to another element (e.g., a second element), it is understood that the element is directly connected to the other element or the element is indirectly connected to the other element via yet another element (e.g., a third element). In contrast, it is understood that when an element (e.g., a first element) is referred to as being "directly connected" or "directly coupled" to another element (a second element), no element (e.g., a third element) is interposed therebetween.
The foregoing description is only exemplary of the preferred embodiments of the invention and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the invention according to the present invention is not limited to the specific combination of the above-mentioned features, but also encompasses other embodiments in which any combination of the above-mentioned features or their equivalents is possible without departing from the scope of the invention as defined by the appended claims. For example, the above features and (but not limited to) features having similar functions disclosed in the present invention are mutually replaced to form the technical solution.

Claims (16)

1. A live broadcast strategy recommendation method comprises the following steps:
acquiring a state index of a live broadcast process node;
and matching the state index of the process node with a preset live broadcast state index, and determining a live broadcast strategy of the process node.
2. The method of claim 1, further comprising:
and recommending live action candidates according to the live strategy of the process node.
3. The method of claim 2, further comprising:
determining a selected action in the live action candidates;
determining a state index contribution value of the selected action to the process node;
and updating the live action candidate according to the contribution value.
4. The method of claim 3, wherein the determining the selected action of the live action candidates comprises:
determining a selected action in a set of live action candidate sets, wherein the set of live action candidate sets comprises at least one live action candidate of red pack, prize drawing, on-shelf merchandise, special coupon, attention placement small card, shop card placement, mouth-broadcast guide attention, and mouth-broadcast question-answer interaction.
5. The method of claim 4, wherein the determining the selected action in the set of live action candidates comprises:
determining the selected action by monitoring each live action candidate in the set of live action candidates.
6. The method of claim 3, wherein the determining the selected action of the live action candidates comprises:
and determining the selected action in the live action candidates by performing multi-mode recognition on the target operation.
7. The method of claim 3, wherein the determining the state metric contribution value of the selected action to the process node comprises:
determining a group of state index dimensions, wherein the group of state index dimensions comprises at least one of average online time, number of newly-added fans, store introduction guidance and total transaction amount;
determining a state indicator contribution value of the selected action to the process node in the set of state indicator dimensions.
8. The method of claims 1-3, further comprising:
establishing a live broadcast state index pool, wherein the live broadcast state index pool comprises a plurality of live broadcast state indexes;
the live broadcast status indexes are as follows: the number of online people, the online time, the number of praise, the number of stored vermicelli, the number of newly added vermicelli, the number of times of guiding to add a shopping cart, the number of collection people and the amount of finished deals.
9. The method of claim 8, wherein the preset live status indicator is determined by:
determining a preset live broadcast state index according to the selection of a user in the live broadcast state index pool; the preset live broadcast state index is one or more.
10. A policy recommendation method comprising:
acquiring a state index of a recorded broadcast process node;
and matching the state index of the process node with a preset live broadcast state index, and determining the interaction strategy of the process node.
11. An interaction method, comprising:
determining a status index of a live process node, wherein the status index of the process node is used for being matched with a preset live state index so as to determine a live strategy of the process node;
and acquiring a live broadcast strategy of the process node.
12. A live policy recommendation apparatus comprising:
the state index acquisition module is used for acquiring the state index of the live broadcast process node;
and the live broadcast strategy determining module is used for matching the state index of the process node with a preset live broadcast state index and determining the live broadcast strategy of the process node.
13. A policy recommendation apparatus comprising:
the state index acquisition module is used for acquiring the state index of the recorded and broadcast process node;
and the strategy determining module is used for matching the state index of the process node with a preset live broadcast state index and determining the interaction strategy of the process node.
14. An interactive device, comprising:
the system comprises a status index determining module, a live broadcast strategy determining module and a live broadcast strategy determining module, wherein the status index of a live broadcast process node is determined and is used for being matched with a preset live broadcast status index so as to determine a live broadcast strategy of the process node;
and the live broadcast strategy acquisition module is used for acquiring the live broadcast strategy of the process node.
15. An electronic device, the device comprising:
the system comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete mutual communication through the communication bus;
the memory is used for storing at least one executable instruction which causes the processor to execute the corresponding operation of the method according to any one of claims 1-11.
16. A storage medium having stored thereon a computer program which, when executed by a processor, carries out the method of any one of claims 1 to 11.
CN202011605791.2A 2020-12-29 2020-12-29 Live broadcast strategy recommendation method, interaction method, device, equipment and storage medium Pending CN114697688A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011605791.2A CN114697688A (en) 2020-12-29 2020-12-29 Live broadcast strategy recommendation method, interaction method, device, equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011605791.2A CN114697688A (en) 2020-12-29 2020-12-29 Live broadcast strategy recommendation method, interaction method, device, equipment and storage medium

Publications (1)

Publication Number Publication Date
CN114697688A true CN114697688A (en) 2022-07-01

Family

ID=82131735

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011605791.2A Pending CN114697688A (en) 2020-12-29 2020-12-29 Live broadcast strategy recommendation method, interaction method, device, equipment and storage medium

Country Status (1)

Country Link
CN (1) CN114697688A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115209174A (en) * 2022-07-18 2022-10-18 忆月启函(盐城)科技有限公司 Audio processing method and system
CN115278286A (en) * 2022-08-02 2022-11-01 抖音视界有限公司 Information processing method and device

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6236980B1 (en) * 1998-04-09 2001-05-22 John P Reese Magazine, online, and broadcast summary recommendation reporting system to aid in decision making
US20110088051A1 (en) * 2008-06-24 2011-04-14 Huawei Technologies Co., Ltd. Method, Apparatus and System for Recommending Media Content
CN105916045A (en) * 2016-05-11 2016-08-31 乐视控股（北京）有限公司 Interactive live broadcast method and device
WO2018188347A1 (en) * 2017-04-11 2018-10-18 武汉斗鱼网络科技有限公司 Information interaction method and apparatus
CN108683927A (en) * 2018-05-07 2018-10-19 广州酷狗计算机科技有限公司 Main broadcaster recommends method, apparatus and storage medium
WO2019071859A1 (en) * 2017-10-09 2019-04-18 武汉斗鱼网络科技有限公司 Method and apparatus for preventing gift swiping on live broadcast platform
CN110401844A (en) * 2019-08-22 2019-11-01 北京字节跳动网络技术有限公司 Generation method, device, equipment and the readable medium of net cast strategy
CN110446057A (en) * 2019-08-30 2019-11-12 北京字节跳动网络技术有限公司 Providing method, device, equipment and the readable medium of auxiliary data is broadcast live
CN110856003A (en) * 2019-10-17 2020-02-28 网易（杭州）网络有限公司 Live list pushing method and device, electronic equipment and storage medium
CN111314719A (en) * 2020-01-22 2020-06-19 北京达佳互联信息技术有限公司 Live broadcast auxiliary method and device, electronic equipment and storage medium
CN111970525A (en) * 2020-08-14 2020-11-20 北京达佳互联信息技术有限公司 Live broadcast room searching method and device, server and storage medium

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6236980B1 (en) * 1998-04-09 2001-05-22 John P Reese Magazine, online, and broadcast summary recommendation reporting system to aid in decision making
US20110088051A1 (en) * 2008-06-24 2011-04-14 Huawei Technologies Co., Ltd. Method, Apparatus and System for Recommending Media Content
CN105916045A (en) * 2016-05-11 2016-08-31 乐视控股（北京）有限公司 Interactive live broadcast method and device
WO2018188347A1 (en) * 2017-04-11 2018-10-18 武汉斗鱼网络科技有限公司 Information interaction method and apparatus
WO2019071859A1 (en) * 2017-10-09 2019-04-18 武汉斗鱼网络科技有限公司 Method and apparatus for preventing gift swiping on live broadcast platform
CN108683927A (en) * 2018-05-07 2018-10-19 广州酷狗计算机科技有限公司 Main broadcaster recommends method, apparatus and storage medium
CN110401844A (en) * 2019-08-22 2019-11-01 北京字节跳动网络技术有限公司 Generation method, device, equipment and the readable medium of net cast strategy
CN110446057A (en) * 2019-08-30 2019-11-12 北京字节跳动网络技术有限公司 Providing method, device, equipment and the readable medium of auxiliary data is broadcast live
CN110856003A (en) * 2019-10-17 2020-02-28 网易（杭州）网络有限公司 Live list pushing method and device, electronic equipment and storage medium
CN111314719A (en) * 2020-01-22 2020-06-19 北京达佳互联信息技术有限公司 Live broadcast auxiliary method and device, electronic equipment and storage medium
CN111970525A (en) * 2020-08-14 2020-11-20 北京达佳互联信息技术有限公司 Live broadcast room searching method and device, server and storage medium

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
刘维付;: "新媒介视域下出版机构淘宝直播营销策略探究", 中国出版, no. 07, 8 April 2020 (2020-04-08) *
张科;: "网络直播的内容生产逻辑及优化策略", 中国编辑, no. 10, 10 October 2020 (2020-10-10) *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115209174A (en) * 2022-07-18 2022-10-18 忆月启函(盐城)科技有限公司 Audio processing method and system
CN115209174B (en) * 2022-07-18 2023-12-01 深圳时代鑫华科技有限公司 Audio processing method and system
CN115278286A (en) * 2022-08-02 2022-11-01 抖音视界有限公司 Information processing method and device

Similar Documents

Publication Publication Date Title
US11360655B2 (en) System and method of non-linear probabilistic forecasting to foster amplified collective intelligence of networked human groups
CN110339569B (en) Method and device for controlling virtual role in game scene
US20100081120A1 (en) Automated quiz generation system
US20220276775A1 (en) System and method for enhanced collaborative forecasting
CN111738294B (en) AI model training method, AI model using method, computer device, and storage medium
CN111760295B (en) Information processing method, information processing device, computer readable medium and electronic equipment
CN114697688A (en) Live broadcast strategy recommendation method, interaction method, device, equipment and storage medium
CN113535991B (en) Multimedia resource recommendation method and device, electronic equipment and storage medium
CN108579095A (en) Social networks in game recommend method, apparatus and computer readable storage medium
CN111701240B (en) Virtual article prompting method and device, storage medium and electronic device
US20220068158A1 (en) Systems and methods to provide mental distress therapy through subject interaction with an interactive space
CN110309753A (en) A kind of race process method of discrimination, device and computer equipment
CN112245936B (en) Account matching method, account matching device and server
CN115375808A (en) Behavior control method and device for virtual live character
Babaei et al. A state-based game attention model for cloud gaming
CN113727762A (en) Video game guidance system
CN117018635A (en) Data processing method, device, electronic equipment and storage medium
US20160379510A1 (en) System and method for adjusting the difficulty of a computer-implemented quiz
Chopade et al. Human-Agent assessment: Interaction and sub-skills scoring for collaborative problem solving
CN113536103B (en) Information recommendation method and device, electronic equipment and storage medium
CN113626681B (en) Account information processing method and device, electronic equipment and storage medium
KR102089520B1 (en) Method for providing sports score predicting pick service considering characteristic variable of possibility of change
US20190385477A1 (en) System and method for sports social media application
CN112819500B (en) Information display method, strategy network training equipment and storage medium
CN115944921B (en) Game data processing method, device, equipment and medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination