CN113590942A - Automatic short video recommendation method and system - Google Patents
Automatic short video recommendation method and system Download PDFInfo
- Publication number
- CN113590942A CN113590942A CN202110811538.0A CN202110811538A CN113590942A CN 113590942 A CN113590942 A CN 113590942A CN 202110811538 A CN202110811538 A CN 202110811538A CN 113590942 A CN113590942 A CN 113590942A
- Authority
- CN
- China
- Prior art keywords
- video
- user
- event information
- videos
- data
- 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
Links
Images
Classifications
-
- 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
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- 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
- G06F16/903—Querying
- G06F16/9035—Filtering based on additional data, e.g. user or group profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The invention discloses a short video automatic recommendation method and a system, wherein the method comprises the following steps: analyzing and preprocessing a user activity log, and respectively generating user event information and video event information; extracting user features according to the user event information, and generating a user portrait and preferences by using a neural network depth model based on feature intersection; dynamically calculating a video score according to the video event information, and dynamically grouping and processing video data according to the video score to generate a plurality of hot data sources; and loading a background configuration operation strategy, recalling videos meeting conditions from different data sources according to the portrait and the preference of the user, and outputting the videos to a front end for display. According to the invention, user preferences are learned according to user logs, various video data sources are generated according to a grouping algorithm, videos meeting conditions are recalled from different data sources through a background configuration operation strategy, and automatic operation of a machine is realized.
Description
Technical Field
The invention belongs to the field of artificial intelligence video recommendation, and particularly discloses a short video machine automatic recommendation method and system.
Background
With the popularization of terminal devices represented by mobile phones and televisions, the mobile phones, the televisions and other terminal devices have more and more roles in replacing computers, and people can browse videos from the internet very conveniently through the mobile phones, the televisions and other terminal devices. When the user group browses the video, the behavior data of the user group is recorded for making video recommendation.
In today's intelligent age, there is a tremendous commercial value behind the enormous data traffic generated by the user population for short video enterprises. How to effectively utilize the flow to achieve accurate recommendation is the core of the recommendation system. The traditional recommendation mode can not avoid causing flow waste, needs a large amount of manual operation to change recommended contents and typesetting, and can not explore and utilize the interests and hobbies of users to recommend proper contents for the users in a personalized way.
Disclosure of Invention
In view of this, the invention provides an automatic short video recommendation method and system, which are used for solving the problem that short videos cannot be automatically recommended.
The invention discloses a short video automatic recommendation method in a first aspect, which comprises the following steps:
analyzing and preprocessing a user activity log, and respectively generating user event information and video event information;
extracting user features according to the user event information, and generating a user portrait and preferences by using a neural network depth model based on feature intersection;
dynamically calculating a video score according to the video event information, and dynamically grouping and processing video data according to the video score to generate a plurality of hot data sources;
and loading a background configuration operation strategy, recalling videos meeting conditions from different data sources according to the portrait and the preference of the user, and outputting the videos to a front end for display.
Preferably, the analyzing and preprocessing the user activity log specifically includes:
and acquiring a user log through a log analysis program, and aggregating click, play and exposure information of a video in a time window to generate different side-load event streams.
Preferably, the extracting user features according to the user event information and the generating the user portrait and the preference by using the neural network depth model based on feature intersection specifically include:
extracting user dimension characteristics and video dimension characteristics; the user dimension features include, but are not limited to, region and viewing time period features of the user, and the video dimension features include, but are not limited to, tag, channel, anchor and viewing duration features of the video viewed by the user;
and performing feature intersection on two dimensions of the user and the video by using a double-tower model to obtain the portrait and the preference of the user.
Preferably, the specific way of dynamically calculating the video score according to the video CTR is as follows:
calculating the CTR of the video by taking the exposure, click and play data of the video;
dynamically calculating a video score S according to the video CTR, the creation time createDay of the video and the ranking time rankDay, wherein the calculation rule is as follows:
S＝(T-createDay)*weight_1+(T-rankDay)*weight_2+CTR
wherein, T is a recommendation period, weight _1 and weight _2 are weight parameters specifically set according to specific services, and the weights are reduced as time increases.
Preferably, the dynamic grouping processing of the video data according to the video scores and generating the hot data source specifically include:
arranging the video scores from high to low, and distributing different groups according to the proportion of the video scores in the global top in a grouping manner; the grouping comprises a new video group, a first-level hot pool, a second-level hot pool and a top-level hot pool, and different groups have different exposure ratios;
subdividing different groups according to the channels, anchor broadcasts or labels of the videos;
and storing the packet data source into an accessible storage medium, and storing the packet data source into storage media according to different scales of time respectively to form a real-time hot data source.
Preferably, the background configuration operation strategy comprises configuring how many floors under one TAB page, the number of videos required by each floor, and adding a limiting condition, wherein the limiting condition comprises the number of floors, the diversity of the anchor and the diversity of the channels.
Preferably, the loading the background configuration operation policy, and recalling videos meeting the conditions from different data sources according to the portrait of the user and the preference, specifically includes:
translating the configured operation strategy into a recall rule and a filtering rule through an analysis program;
under the appointed recall rule, recalling the video which is in line with the preference of the user from the data source by using the portrait and the preference of the user; the data sources comprise an offline data source, a hot data source and a recently updated data source, and are recalled from top to bottom in sequence;
reordering the recalled data according to the video score, and outputting the data meeting the requirements according to the filtering rule;
and finally, assembling all the floor output JSON data and returning the JSON data to the caller.
In a second aspect of the present invention, a short video automatic recommendation system is disclosed, the system comprising:
a log analysis module: the system is used for analyzing and preprocessing the user activity log and respectively generating user event information and video event information;
an image generation module: extracting user features from the user event information, generating a user profile and preferences using a neural network depth model based on feature intersection;
a dynamic grouping module: dynamically calculating a video score according to the video event information, and dynamically grouping and processing video data according to the video score to generate a plurality of hot data sources;
the data recall module: and loading a background configuration operation strategy, recalling videos meeting conditions from different data sources according to the portrait and the preference of the user, and outputting the videos to a front end for display.
In a third aspect of the present invention, an electronic device is disclosed, comprising: at least one processor, at least one memory, a communication interface, and a bus;
the processor, the memory and the communication interface complete mutual communication through the bus;
the memory stores program instructions executable by the processor which are invoked by the processor to implement the method of the first aspect of the invention.
In a fourth aspect of the invention, a computer-readable storage medium is disclosed, which stores computer instructions for causing a computer to implement the method of the first aspect of the invention.
Compared with the prior art, the invention has the following beneficial effects:
1) extracting user characteristics according to a user log, generating user portrait and preference by using a neural network depth model based on characteristic intersection, dynamically calculating video scores according to video event information by deeply mining user preference, and dynamically grouping video data according to the video scores to generate a plurality of hot data sources, so that the personalized recommendation of a user is realized, and simultaneously, the balance between utilization and exploration is realized by combining a video dynamic grouping algorithm, and the user viscosity is improved;
2) and (3) the background is configured with an operation strategy to generate a recall rule and a filtering rule of data, videos meeting conditions are recalled from different data sources according to images and preferences of users, all floors are assembled, automatic operation of a machine is realized, automatic typesetting and weight discharging are realized, and manual operation is reduced.
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 of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a schematic flow chart of a short video automatic recommendation method according to the present invention;
fig. 2 is a schematic structural diagram of the short video automatic recommendation system of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to 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 of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
Referring to fig. 1, a flow chart of the short video automatic recommendation method of the present invention is schematically illustrated; the invention provides a short video automatic recommendation method which comprises the following steps:
s1, analyzing and preprocessing the user activity log, and respectively generating user event information and video event information;
setting a server log address to be acquired, monitoring the server log in real time by installing a log analysis program, and acquiring data of the server log when the log starts to generate a new entry and sending the data into the log analysis program.
And extracting video information in the log obtained by the log analysis program, and removing other information impurities so as to collect the history record and browsing behavior of the video watched by the user. And aggregating click, play, exposure and other information of one video in a time window to generate different event streams of the emphasis. Such as: user information time flow, information flow of exposure, click, playing and the like of a certain video of a certain floor of a certain user under a certain TAB of a certain channel, user event flow is generated by aggregating the viewing history of the user according to the user and connecting the context information in series, and finally user event information is generated. And similarly, video event information is generated according to the information of clicking, playing, exposing and the like of the video.
S2, extracting user characteristics according to the user event information, and generating a user portrait and preference by using a neural network depth model based on characteristic intersection;
extracting user dimension characteristics and video dimension characteristics from the user event information; the user dimension characteristics comprise characteristics of regions, watching time periods and the like of users, and the video dimension characteristics comprise characteristics of tags, channels, anchor broadcasts, watching time periods and the like of videos watched by the users;
and performing feature intersection on two dimensions of the user and the video by using a double-tower model to obtain the portrait of the user and the preference, wherein the preference comprises channel, anchor and label preference.
S3, dynamically calculating video scores according to the video event information, and dynamically grouping the video data according to the video scores to generate a plurality of hot data sources; the method comprises the following steps of generating a hot door data source through a grouping algorithm, wherein the grouping algorithm is an algorithm for distinguishing the quality of videos, and the videos are classified from good to bad.
Acquiring an activity log of a user in a server;
extracting video information in the video information, and removing other information impurities;
aggregating data of video exposure, Click, play and the like according to a time window, and calculating a video CTR (Click Through Rate) according to the data of video exposure, Click and play;
then, grouping the data according to multiple dimensions such as the creation time, the CTR, the top list retention time and the like of the video, specifically, dynamically calculating a video score S according to the CTR of the video, the creation time createDay of the video and the top list time rankDay, wherein the calculation rule is as follows:
S＝(T-createDay)*weight_1+(T-rankDay)*weight_2+CTR
wherein, T is a recommendation period, weight _1 and weight _2 are weight parameters specifically set according to specific services, and the weights are reduced as time increases.
The way in which the video scores are calculated is dynamic, changing the mobility of the videos within different groups based on the user's interaction behavior. Arranging the video scores from high to low, and distributing different groups according to the proportion of the video scores in the global top in a grouping manner; the groups generally comprise a new video group, a second-level hot pool, a first-level hot pool and a top-level hot pool, different groups have different exposure ratios, the exposure ratios are defined by the groups, and the higher the ratio is, the more the video recommendation times are. For example, after each video score is estimated, the video scores are arranged from high to low, if the estimated video score of a certain video is ranked at the top and belongs to the top 10% (top 10%) of the global ranking, the video score can be divided into a top hot pool, if the estimated video score of another video belongs to the top 10% -30%, the video score is divided into a first hot pool, and so on. And the different groups are subdivided according to the channels, the anchor broadcasts, the labels and the like of the videos.
The packet data source is stored in an accessible storage medium and is respectively stored according to different scales of time such as hours, days and the like, so that a real-time hot data source is formed.
According to the method, the user characteristics are extracted according to the user log, the neural network depth model based on characteristic intersection is used for generating the user portrait and the preference, the user preference is deeply mined, the video score is dynamically calculated according to the video event information, the video data are dynamically grouped according to the video score to generate a plurality of hot data sources, personalized recommendation of the user is achieved, balance between utilization and exploration is achieved by combining a video dynamic grouping algorithm, and the user viscosity is improved.
And S4, loading a background configuration operation strategy, recalling videos meeting conditions from different data sources according to the portrait and the preference of the user, and outputting the videos to a front end for display.
The machine automation operation effect is realized by configuring an operation strategy, such as: how many floors there are under a TAB page, the number of videos required per floor, and restrictions such as channel restrictions, anchor restrictions, or diversity restrictions are added. Then, translating the configured operation strategy into a recall rule and a filtering rule through an analysis program, recalling videos such as channels, anchor broadcasts, labels and the like which accord with the preference of a user from a data source by utilizing the portrait and the preference of the user under the appointed recall rule, wherein the data source comprises an offline data source, a hot data source and a recently updated data source, recalling in sequence according to the sequence of the offline data source, the hot data source and the recently updated data source, sequencing the data in each data source according to a CTR (program counter) sequence, recalling in sequence from top to bottom in each data source, reordering the recalled data according to the video score, and outputting the data which accord with the requirement according to the filtering rule; and finally, assembling all floors, filling data into the structural body of the specified floor, outputting JSON data to return to a caller, and recommending videos meeting the user preference.
The method and the system have the advantages that the recall rule and the filtering rule of the generated data are analyzed by loading the background configuration operation strategy, the videos meeting the conditions are recalled from different data sources by combining the portrait and the preference of the user, all floors are assembled, the automatic operation of a machine is realized, the automatic typesetting and the rearrangement are realized, and the manual operation is reduced.
Corresponding to the embodiment of the method, the invention also discloses a short video automatic recommendation system, please refer to fig. 2, the system includes:
the log analysis module 10: the system is used for analyzing and preprocessing the user activity log and respectively generating user event information and video event information;
the portrait generation module 20: extracting user features from the user event information, generating a user profile and preferences using a neural network depth model based on feature intersection;
dynamic grouping module 30: dynamically calculating a video score according to the video event information, and dynamically grouping and processing video data according to the video score to generate a plurality of hot data sources;
the data recall module 40: and loading a background configuration operation strategy, recalling videos meeting conditions from different data sources according to the portrait and the preference of the user, and outputting the videos to a front end for display.
The above system embodiments are suitable for the method embodiments in a one-to-one correspondence, and please refer to the method embodiments for brief description of the system embodiments.
According to the method, historical records and browsing behaviors of videos watched by a user are collected, preprocessing operation is performed on the data, the processed data are input into a training model, and the portrait of the user and the preference of the user are predicted; collecting data of video exposure, clicking, playing and the like, inputting the data into a grouping algorithm to estimate video scores, grouping all videos according to the video scores, wherein different groups have different exposure granularities, the higher the video score is, the higher the exposure possibility is, and the weight is increased or attenuated along with the data of exposure, clicking, playing and the like; finally, recalling conditions of different floors of different plates is appointed through a recommendation background unified configuration frame, and videos meeting the conditions are recalled from different grouped data sources according to portraits and preferences of users and output to a front end for display. According to the invention, the user preference is deeply mined, the personalized recommendation of the user is realized, on one hand, the balance of utilization and exploration is realized, the viscosity of the user is improved, on the other hand, the automatic operation of a machine is realized, the automatic typesetting and the rearrangement can be realized, the manual operation is reduced, and the method can be applied to the automatic recommendation of terminal equipment such as a mobile phone terminal, a television terminal, a computer terminal or a tablet personal computer.
The present invention also discloses an electronic device, comprising: at least one processor, at least one memory, a communication interface, and a bus; the processor, the memory and the communication interface complete mutual communication through the bus; the memory stores program instructions executable by the processor, and the processor calls the program instructions to implement all or part of the steps of the short video automatic recommendation method.
The invention also discloses a computer-readable storage medium which stores computer instructions, and the computer instructions enable the computer to realize all or part of the steps of the automatic short video recommendation method in the embodiment of the invention. The storage medium includes: u disk, removable hard disk, ROM, RAM, magnetic disk or optical disk, etc.
The above-described system embodiments are merely illustrative, wherein the units described as separate parts may or may not be physically separate, and the parts shown as units may or may not be physical units, i.e. may be distributed over a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (10)
1. A short video automated recommendation method, the method comprising:
analyzing and preprocessing a user activity log, and respectively generating user event information and video event information;
extracting user features according to the user event information, and generating a user portrait and preferences by using a neural network depth model based on feature intersection;
dynamically calculating a video score according to the video event information, and dynamically grouping and processing video data according to the video score to generate a plurality of hot data sources;
and loading a background configuration operation strategy, recalling videos meeting conditions from different data sources according to the portrait and the preference of the user, and outputting the videos to a front end for display.
2. The method for automatically recommending short videos according to claim 1, wherein the parsing and preprocessing the user activity log specifically comprises:
and acquiring a user log through a log analysis program, and aggregating click, play and exposure information of a video in a time window to generate different side-load event streams.
3. The method for automatically recommending short videos according to claim 2, wherein the extracting user features according to user event information and the generating user portrait and preference using a neural network depth model based on feature intersection specifically comprises:
extracting user dimension characteristics and video dimension characteristics from the user event information; the user dimension features include, but are not limited to, region and viewing time period features of the user, and the video dimension features include, but are not limited to, tag, channel, anchor and viewing duration features of the video viewed by the user;
and performing feature intersection on two dimensions of the user and the video by using a double-tower model to obtain the portrait and the preference of the user.
4. The method for automatically recommending short videos according to claim 2, wherein the specific way for dynamically calculating the video score according to the video CTR is:
calculating the CTR of the video by taking the exposure, click and play data of the video;
dynamically calculating a video score S according to the video CTR, the creation time createDay of the video and the ranking time rankDay, wherein the calculation rule is as follows:
S＝(T-createDay)*weight_1+(T-rankDay)*weight_2+CTR
wherein, T is a recommendation period, weight _1 and weight _2 are weight parameters specifically set according to specific services, and the weights are reduced as time increases.
5. The method for automatically recommending short videos according to claim 4, wherein the dynamically grouping video data according to video scores and generating hot data sources are specifically:
arranging the video scores from high to low, and distributing different groups according to the proportion of the video scores in the global top; the grouping comprises a new video group, a first-level hot pool, a second-level hot pool and a top-level hot pool, and different groups have different exposure ratios;
subdividing different groups according to the channels, anchor broadcasts or labels of the videos;
and storing the packet data source into an accessible storage medium, and storing the packet data source into storage media according to different scales of time respectively to form a real-time hot data source.
6. The short video automatic recommendation method according to claim 1, wherein the background configuration operation strategy comprises configuring how many floors under a TAB page, the number of videos required by each floor, and adding a limiting condition, wherein the limiting condition comprises the number of floors, the anchor diversity and the channel diversity.
7. The method according to claim 6, wherein the loading of the background configuration operation policy and the recall of eligible videos from different data sources according to the portrait and the preference of the user specifically comprise:
translating the configured operation strategy into a recall rule and a filtering rule through an analysis program;
under the appointed recall rule, recalling the video which is in line with the preference of the user from the data source by using the portrait and the preference of the user; the data sources comprise an offline data source, a hot data source and a recently updated data source, and are recalled from top to bottom in sequence;
reordering the recalled data according to the video score, and outputting the data meeting the requirements according to the filtering rule;
and finally, assembling all the floor output JSON data and returning the JSON data to the caller.
8. A short video automated recommendation system, the system comprising:
a log analysis module: the system is used for analyzing and preprocessing the user activity log and respectively generating user event information and video event information;
an image generation module: extracting user features from the user event information, generating a user profile and preferences using a neural network depth model based on feature intersection;
a dynamic grouping module: dynamically calculating a video score according to the video event information, and dynamically grouping and processing video data according to the video score to generate a plurality of hot data sources;
the data recall module: and loading a background configuration operation strategy, recalling videos meeting conditions from different data sources according to the portrait and the preference of the user, and outputting the videos to a front end for display.
9. An electronic device, comprising: at least one processor, at least one memory, a communication interface, and a bus;
the processor, the memory and the communication interface complete mutual communication through the bus;
the memory stores program instructions executable by the processor, the processor invoking the program instructions to implement the method of any one of claims 1-7.
10. A computer readable storage medium storing computer instructions for causing a computer to implement the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110811538.0A CN113590942A (en) | 2021-07-19 | 2021-07-19 | Automatic short video recommendation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110811538.0A CN113590942A (en) | 2021-07-19 | 2021-07-19 | Automatic short video recommendation method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113590942A true CN113590942A (en) | 2021-11-02 |
Family
ID=78248448
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110811538.0A Pending CN113590942A (en) | 2021-07-19 | 2021-07-19 | Automatic short video recommendation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113590942A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115878841A (en) * | 2023-03-03 | 2023-03-31 | 南京邮电大学 | Short video recommendation method and system based on improved bald eagle search algorithm |
-
2021
- 2021-07-19 CN CN202110811538.0A patent/CN113590942A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115878841A (en) * | 2023-03-03 | 2023-03-31 | 南京邮电大学 | Short video recommendation method and system based on improved bald eagle search algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110781391B (en) | Information recommendation method, device, equipment and storage medium | |
| CN110598016B (en) | Method, device, equipment and medium for recommending multimedia information | |
| CN109241425B (en) | Resource recommendation method, device, equipment and storage medium | |
| CN110781321B (en) | Multimedia content recommendation method and device | |
| CN111382361B (en) | Information pushing method, device, storage medium and computer equipment | |
| CN112241327A (en) | Shared information processing method and device, storage medium and electronic equipment | |
| CN109992715B (en) | Information display method, device, medium and computing equipment | |
| CN109376237A (en) | Prediction technique, device, computer equipment and the storage medium of client's stability | |
| US20200106734A1 (en) | Latent user communities | |
| CN113742567B (en) | Recommendation method and device for multimedia resources, electronic equipment and storage medium | |
| CN109903087A (en) | The method, apparatus and storage medium of Behavior-based control feature prediction user property value | |
| CN113301376B (en) | Live broadcast interaction method and system based on virtual reality technology | |
| CN108133390A (en) | For predicting the method and apparatus of user behavior and computing device | |
| CN112165639A (en) | Content distribution method, content distribution device, electronic equipment and storage medium | |
| CN115757952A (en) | Content information recommendation method, device, equipment and storage medium | |
| CN111552835A (en) | File recommendation method and device and server | |
| CN113297486B (en) | Click rate prediction method and related device | |
| CN113590942A (en) | Automatic short video recommendation method and system | |
| CN113626638A (en) | Short video recommendation processing method and device, intelligent terminal and storage medium | |
| CN115514995A (en) | Method, device and equipment for displaying recommendation information of live broadcast room | |
| CN111064996B (en) | Method, system and storage medium for identifying user watching video content preference | |
| CN110427620B (en) | Service quality optimization management system based on community system | |
| CN112269943A (en) | Information recommendation system and method | |
| CN113297461B (en) | Target user identification method, target user group identification method and device | |
| CN112291625B (en) | Information quality processing method, information quality processing device, electronic equipment and storage 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 |