CN109769128B - Video recommendation method, video recommendation device and computer-readable storage medium - Google Patents
Video recommendation method, video recommendation device and computer-readable storage medium Download PDFInfo
- Publication number
- CN109769128B CN109769128B CN201811588951.XA CN201811588951A CN109769128B CN 109769128 B CN109769128 B CN 109769128B CN 201811588951 A CN201811588951 A CN 201811588951A CN 109769128 B CN109769128 B CN 109769128B
- Authority
- CN
- China
- Prior art keywords
- video
- candidate
- live
- videos
- live broadcast
- 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.)
- Active
Links
Images
Landscapes
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application relates to a video recommendation method, a video recommendation device and a computer readable storage medium. The video recommendation method comprises the following steps: acquiring historical sharing data of a user; generating a candidate video set and a candidate live broadcast set according to the historical sharing data; ranking all candidate videos in the candidate video set; and fusing the sorted candidate video set and the candidate live broadcast set to be recommended as recommended content. According to the video recommendation method, live broadcast recommendation is integrated into short video recommendation, the requirements of a user on viewing experience and content diversity of recommended content are enhanced, and the recommended video can be provided for the user more accurately and comprehensively.
Description
Technical Field
The application belongs to the field of computer software application, and particularly relates to a video recommendation method and a video recommendation device.
Background
The sharing function of APP plays an important role in the transmission and transformation of products. Most of APPs in the market have the function of sharing by a third party, and the user is enabled to actively share the content in the APP to the social platform by means of the flow of social products, so that more users can watch the content in the APP, and the user can watch the content conveniently. Generally, in order to give more content choices to the user, short video content that may be of interest to the user is recommended as much as possible on the shared short video display interface.
However, most of the current recommendation systems mainly recommend individual item information. Taking recommendation of short videos of a sharing page as an example, the likeness degree (such as clicking, praise, comment or share) of a user to the short videos can be found through historical behavior data of the user, the likeness is measured and scored, the relationship among the users is calculated according to attitudes and preference degrees of different users to the same short videos, and then recommendation of similar short videos is carried out among the users with the same likeness.
The inventor finds that the style of the content of the short video sharing page selected by the recommendation method is relatively single, the short video sharing page is not easy to attract users, and the user experience is poor.
Disclosure of Invention
In order to solve the problems in the related art, the application discloses a video recommendation method and a video recommendation device, live videos are fused in short video sharing, recommendation is carried out simultaneously, recommendation diversity of the short videos is improved, and user experience is improved.
According to a first aspect of embodiments of the present application, there is provided a video recommendation method, including:
acquiring historical sharing data of a user;
generating a candidate video set and a candidate live broadcast set according to the historical sharing data;
ranking all candidate videos in the candidate video set; and
and fusing the sorted candidate video set and the candidate live broadcast set to be recommended as recommended content.
Optionally, the video recommendation method further includes: and sending the recommended content to a user side for displaying.
Optionally, the history sharing data includes history sharing content and a degree of attention obtained by the history sharing content.
Optionally, the history sharing content includes a history sharing video and an on-air live video, and a part of the history sharing video and a part of the on-air live video are added to the candidate video set and the candidate live video set respectively according to a preset rule.
Optionally, the video recommendation method further includes: and auditing all on-broadcast live videos in the candidate live broadcast set, and removing the on-broadcast live videos which are not approved.
Optionally, an audit time is set, an audit state of each live video is monitored, and if an instruction that the audit is passed is not received within the audit time, the live video is deleted from the candidate live broadcast set.
Optionally, generating a candidate live broadcast set according to the historical sharing data includes:
inquiring live broadcast IDs and affiliated anchor IDs of all live broadcast videos at the current moment;
counting attention of each live broadcast video; and
and adding the part of the on-air live video with higher attention into the candidate live video set.
Optionally, generating a candidate live broadcast set according to the historical sharing data further includes:
obtaining friend IDs with similar relations to the anchor IDs;
and adding part of the friend ID into the candidate live broadcast set in the live broadcast video.
Optionally, the recommending, as recommended content, the merged ranked candidate video set and candidate live broadcast set includes:
acquiring a video list of all the candidate videos after sequencing;
sequencing the live broadcast videos to obtain a live broadcast list;
and the live broadcast list is merged into the video list to be used as the recommended content for recommendation.
According to a second aspect of the embodiments of the present invention, there is provided a video recommendation apparatus including:
the acquisition module is used for acquiring historical shared data of a user;
the candidate module is used for generating a candidate video set and a candidate live broadcast set according to the historical sharing data;
the sorting module is used for sorting all the candidate videos in the candidate video set; and
and the recommending module is used for fusing the sequenced candidate video set and the candidate live broadcast set to recommend the candidate video set and the candidate live broadcast set as recommended content.
Optionally, the history sharing data includes history sharing content and a degree of attention obtained by the history sharing content.
Optionally, the history shared content includes a history shared video and an on-air live video, and a part of the history shared video and a part of the on-air live video are selected and added into the candidate video set and the candidate live set respectively.
Optionally, the video recommendation apparatus further includes: and the auditing module is used for auditing all live broadcast videos in the candidate live broadcast set and removing the live broadcast videos which are not approved.
Optionally, an audit time is set, an audit state of each live video is monitored, and if an instruction that the audit is passed is not received within the audit time, the live video is deleted from the candidate live broadcast set.
Optionally, the candidate module comprises:
the query module is used for querying the live broadcast IDs and the affiliated anchor IDs of all live broadcast videos at the current moment;
the statistic module is used for counting the attention of each piece of live broadcast video;
and the first selection module is used for adding the part of the on-air live video with higher attention into the candidate live video set.
Optionally, the candidate module further comprises:
a friend relation obtaining module, configured to obtain a friend ID having a similar relation to the anchor ID; and
and the second selection module is used for adding part of the friend ID into the candidate live broadcast set in the live broadcast video.
Optionally, the recommendation module includes:
the video list acquisition module is used for acquiring a video list of all the candidate videos after sequencing;
the live broadcast list acquisition module is used for sequencing the live broadcast video to obtain a live broadcast list;
and the fusion module is used for fusing the live list into the video list to serve as the recommended content for recommendation.
According to a third aspect of an embodiment of the present invention, there is provided an electronic apparatus, including:
a processor;
a memory for storing processor-executable instructions;
wherein the processor is configured to perform any one of the video recommendation methods described above.
According to a fourth aspect of the embodiments of the present invention, there is provided a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions, when executed, implement the above-mentioned video recommendation method.
The technical scheme provided by the embodiment of the application can have the following beneficial effects:
according to the video recommendation method, the historical sharing data of the user is obtained, the candidate video set and the candidate live broadcast set are generated according to the historical sharing data, the candidate video set and the candidate live broadcast set are fused and then recommended as recommended contents, live broadcast contents are added in the short video sharing, the video recommendation diversity of a sharing page is increased, a live broadcast form is added in the short video sharing, the interest of the user can be aroused, and the watching experience of the user is improved.
The technical scheme provided by another embodiment of the application can have the following beneficial effects:
by sequencing all candidate videos in the candidate video set, sequencing the videos which are broadcast and live broadcast and sequentially recommending according to the sequenced result, videos which are interested by most people and live broadcast can be shared simultaneously, the requirements of most users are met, the recommendation quality of the videos is guaranteed, the watching experience of the users is improved, and the interest of the users is promoted.
The technical scheme provided by another embodiment of the application can have the following beneficial effects:
in order to guarantee the quality of the recommended live video, auditing time is set, all live videos in the candidate live broadcast set are audited, the live videos which do not pass the auditing are removed, and some contents with high real-time heat are recommended to the user, so that the video recommendation is more suitable for the user preference, and the recommendation quality is improved.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
FIG. 1 is a flow diagram illustrating a video recommendation method in accordance with an exemplary embodiment;
FIG. 2 is a flow diagram illustrating a summarized video recommendation method in accordance with an illustrative embodiment;
fig. 3 is a partial flowchart according to an exemplary embodiment of step S202 in the video recommendation method shown in fig. 2;
fig. 4 is a flowchart according to an exemplary embodiment of step S205 in the video recommendation method shown in fig. 2;
FIG. 5 is a schematic diagram of a video recommendation device shown in accordance with an exemplary embodiment;
FIG. 6 is a schematic diagram of an aggregated video recommender shown in accordance with an exemplary embodiment;
FIG. 7 is a schematic diagram of an aggregated video recommender shown in accordance with an exemplary embodiment;
FIG. 8 is a block diagram illustrating an electronic device performing a video recommendation method in accordance with an exemplary embodiment;
fig. 9 is a block diagram illustrating a video recommendation apparatus that performs a video recommendation method according to an example embodiment.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.
Fig. 1 is a flowchart illustrating a video recommendation method according to an exemplary embodiment, which specifically includes the following steps.
In step S101, history sharing data of the user is acquired.
As mentioned in the background art, the sharing content of the short video sharing page selected by the current video recommendation method is relatively single in style, is not easy to attract users, and is poor in experience, so that the application provides a new video recommendation method, live broadcast results are fused in short video sharing, the diversity of video recommendation is improved, and the user interest is increased.
Generally, a user browses videos inside an APP, when finding an interested short video envelope, the user clicks to enter the viewing, and on the way of viewing the short video, a series of operations such as praise (representing a favorite), comment or reply (issuing an interaction), forwarding (recommending to others) and the like may be performed. When a user views a particularly favorite short video or live video, the user usually shares the video through various social platforms (such as WeChat, a friend group, and the like), and the shared video can be viewed by other users to perform operations such as approval, viewing, comment, collection, and the like.
The server receives the data request, records all operations of the user in the APP and all operated records of the short video or the live video in the shared page or in the APP, and finally stores the records in a log form, and can extract log data when needed.
In this embodiment, to perform a fine and reasonable video recommendation, historical shared data is needed, so the historical shared data of the user is collected and log data of the user in a period of time is extracted.
In one embodiment, the history sharing data includes history sharing content and attention gained by the history sharing content. As described above, the user may perform corresponding operations when seeing the shared content shared on the shared page, and these operations on the history shared content are regarded as attention degrees, which are, for example, the number of clicks, the number of comments, the viewing time, and the number of forwarding times (when the user views the content shared by a plurality of friends on the shared page of the social platform, the times of clicks, views, and the like of different shared contents are different due to different interests and hobbies), … …, and the comprehensive evaluation is, for example, to assign a weight value to each item, then to convert the index values of each item into scores, to score the attention degrees, and to determine the quality of the history shared content according to the final attention degree score.
When a user watches the shared short videos or live videos on the sharing page of the social platform, if videos which are interesting are available, the user can click to check the videos, and even the videos enter the APP to watch the videos so as to know more wonderful contents, so that the watching interest of the user is improved, the user can better accord with the preference of the user, the user can independently select the watching type which is liked by the user, and the watching experience is improved.
In step S102, a candidate video set and a candidate live broadcast set are generated according to the history sharing data.
The core idea of the embodiment is that sharing of some live videos is added while short videos are recommended to a user, and diversity and comprehensiveness of recommendation are improved, so that the short videos and the live videos which are used as recommended contents need to be selected, the selected short videos are sorted together to serve as a candidate video set, and the selected live videos are sorted together to serve as a candidate live set. As can be seen from step S101, since the history shared data of the user can be extracted from the server if necessary, in this step, the candidate video set and the candidate live broadcast set are selected based on the history shared data.
In one embodiment, the history shared content includes a history shared video and a live video, the history shared video is a short video shared by the user, and the live video is a live video being played. And selecting short videos and live videos suitable for recommendation according to the attention degree of the historical shared videos and the on-broadcasting live videos.
Firstly, collecting the shared historical shared videos and all live broadcast videos which are in broadcasting, and recalling recommended contents from a large number of short videos and live broadcast videos according to a certain rule, namely adding part of the historical shared videos and part of the live broadcast videos into a candidate video set and a candidate live broadcast set respectively.
For example, recommended content is selected according to the attention degree obtained from the history shared content, so the attention degrees of the recalled short videos and live videos are respectively sorted, an attention degree threshold is set, and the short videos and the live videos larger than the attention degree threshold are added to the recommended content. The attention of the live video can be regarded as the real-time heat of the live video, and the attention of the short video can be regarded as a measure index of the high and low video quality, and can also be understood as a preset rule.
In one embodiment, when video recall is performed, screening is performed first, videos without click rate or with click rate lower than a preset lower limit are removed, and then attention degrees of other videos are calculated, so that a large amount of time and calculation amount can be saved, and screening efficiency is improved.
In one embodiment, a few short videos with high attention are analyzed, for example, 30 history shared videos with the highest attention are selected, a part of videos with high similarity to the selected videos is selected from the APP, and the selected videos are also added into the candidate video set.
Personalized recommendations may reference the ideas of "Match" and "Rank" in computational advertisers: "Match" is an effective and rich recall that finds content positively correlated to the user from the total content as much as possible and returns the result to "Rank". For example, 1000 videos most relevant to the user interest are preliminarily selected from billions of videos to serve as a candidate set finally recommended to the user. This is a rough and rapid screening process, and the screening method is simple and is generally called as "rough screening".
And the Rank stage is used for sorting the contents returned by the Match stage, generally called as a sorting stage, predicting the possibility of the user clicking the contents, and recommending the contents which are most likely to be clicked to the user firstly. Since the model of the Rank stage is complex and it is impossible to sort all the contents, it is necessary to roughly screen the data through the Match stage. And the data screened by the Match directly determines the data quality of the Rank stage, and the recommended result is greatly influenced.
The steps are mainly coarse screening in the 'Match' stage, collected short videos and live videos are used as a candidate video set and a candidate live broadcast set in the 'Match' stage, and the following steps are carried out on recommended contents in the 'Rank' stage and then displayed to a user after the recommended contents are processed and sorted.
In step S103, all candidate videos in the candidate video set are sorted.
The step is to finely select the candidate video set, and the selected short video is used as the final recommended content.
All candidate videos in the recalled candidate video set are ranked, and some ranking algorithms can be adopted to rank the candidate videos according to the attention degree score or the video size and the like. For example, the short videos are sorted according to the number of times of being praised of each short video, the top N short videos are extracted for recommendation, and the rest short videos are deleted from the candidate video set. For example, the short videos with too long playing time are sorted according to the playing time of the short videos, and the short videos with too long playing time are deleted from the candidate video set, so that the watching time is saved, and the sharing efficiency is improved.
According to the steps, the attention degree of each short video and each live video is scored according to historical sharing content on a sharing page and operation behaviors of a user, the short videos and the live videos are ranked according to the attention degree, part of historical sharing videos with high attention degree are selected from a recalled video library to be added into a candidate video set, and part of live videos with high live broadcast heat degree are selected from all current live videos to be added into the candidate live broadcast set.
For example, all short videos and live videos which are shared from an APP to a certain social platform within a certain time are collected, whether the attention degree of the videos reaches the attention degree threshold value is calculated respectively, videos which do not reach the attention degree threshold value are removed, and videos with poor quality can be prevented from being recommended to users. For example, videos published in a past month of a certain social platform sharing page are collected, the number of likes and comments accumulated in the past 15 days of the videos is calculated, the interest degree obtained by the videos is represented by the number of the likes or clicks or the number of the comments of the user on the videos, an interest degree threshold value is set, for example, 10000 clicks are accumulated, videos with the interest degree not reaching the threshold value are filtered, and videos with the interest degree higher than the threshold value can be added into a candidate video set as videos to be recommended. Therefore, videos with low attention can be removed, and videos which are most of users interested in are selected for recommendation, so that the quality of recommended videos is guaranteed.
In one embodiment, the ranked candidate videos in the candidate video set are reviewed, for example, the content of the short videos is manually reviewed, and the short videos meeting the current situation and popularity are selected for recommendation, so that the recommended videos have epoch colors, the watching desire of the user is improved, and the conversion rate of the user is improved.
In one embodiment, all candidate videos in the candidate video set are sorted and then ranked, for example, all short videos are divided into a fun class, an emotion class, a moral class, a music class, etc., each video is ranked according to the attention degree, and then ranked according to the fun class, the music class, the emotion class, the moral class … ….
After sequencing, the remaining videos in the candidate video set are high-quality videos after fine screening, and a very high click rate and a very high conversion rate are obtained when sharing is predicted.
In step S104, the sorted candidate video set and the candidate live broadcast set are merged and recommended as recommended content.
When a user watches short videos inside an APP or on a shared page, after a short video is watched, the user can normally automatically play the next short video, or when watching live videos shared by other users, the user can turn over the video upwards or downwards or left and right, so that other live videos can be watched, and the positions of the short videos and the live videos are stored to serve as a recommendation pool. In this embodiment, the recommendation pool for the short videos and the recommendation pool for the live videos are shared, that is, the recommendation pool can display the recommended short videos and also can display the recommended live videos, so that the richness of recommended contents in the recommendation pool is increased.
In step S102, a candidate video set and a candidate live broadcast set, which are respectively composed of a part of short videos with a higher attention and a part of live broadcasts with a higher popularity, are obtained, and in this step, the candidate video set and the candidate live broadcast set are fused, that is, candidate videos in the ordered candidate video set (the candidate videos are short videos to be recommended) and live broadcasts in the candidate live broadcast set are fused, and video recommendation is performed according to the fused sequence.
When the short videos and the live videos are fused, the fusion and recommendation can be performed according to different numbers (for example, the number of the live videos is larger than that of the short videos) and different sequences (the live videos are recommended first and then or the live videos and the live videos are recommended alternately), and then the fused recommendation content is added into a recommendation pool.
In one embodiment, when the user sends or shares the recommended content, the user submits a request for sending the recommended content to the server, and the server receives the request and sends the video content.
In the embodiment, through obtaining historical sharing data of a user, a candidate video set and a candidate live broadcast set are generated according to the historical sharing data, and then the candidate video set and the candidate live broadcast set are fused and recommended as recommended contents, live broadcast contents are added in video sharing, the update and live broadcast conversion rates of users sharing a page are increased, a live broadcast form is added, the users are more willing to enter APP internal activities, and the user conversion rate in short-video social sharing is improved.
FIG. 2 is a flow diagram illustrating a summarized video recommendation method according to an example embodiment.
In step S201, historical shared data of a user is acquired;
in step S202, a candidate video set and a candidate live broadcast set are generated according to the historical sharing data;
in step S203, all candidate videos in the candidate video set are sorted;
in step S204, all live videos in the candidate live broadcast set are audited, and live videos in live broadcast that do not pass the audition are removed;
in step S205, the candidate video set and the candidate live broadcast set are merged and recommended as recommended content.
This embodiment is the optimization scheme of fig. 1, and steps S201 to S203 and step S205 are the same as steps S101 to S104 of fig. 1, and are not described again here.
In the first embodiment, recall of recommended content is mainly performed, a high-quality video is to be recommended to a user, and fusion is performed after fine screening is performed on candidate videos and live videos respectively, as shown in step S204.
In step S204, all live videos in the candidate live broadcast set are reviewed, and live videos in live broadcasts that do not pass the review are removed.
In this step, all live videos in the candidate live broadcast set are finely screened, for example, all live videos in the live broadcast are audited and filtered, and live videos which do not pass the audition are deleted according to the audition result. The auditing is for example auditing the safety and the legality of the live video, filtering out illegal videos, ensuring the safety and the health of a network environment, and recommending the live video with green and positive energy to a user.
And uploading the live video during auditing, detecting the legality of the live video by using a detection algorithm, and feeding back an auditing result indication after a certain time, for example, an auditing passing message, so that the live video can be used as the live video to be recommended.
In one embodiment, a Deep Packet Inspection (DPI) is adopted to perform auditing of live broadcast videos, the DPI is a Packet-based deep Packet inspection technology, the DPI performs deep inspection on different network application layer loads (such as HTTP, DNS and the like), the validity of the DPI is determined by detecting payloads of the DPI, and after the DPI is detected, illegal live broadcast videos are prohibited from being broadcast and warned, and the DPI is deleted from a candidate video set.
In an embodiment, an audit time (e.g., T) is set, an audit state of each piece of live video in the broadcast is monitored, and if an instruction that the audit is passed is not received within the audit time, that is, an indication that the audit is passed is not yet after the audit time T is exceeded, indicating that the live video cannot be pushed, the live video in the broadcast is deleted from the candidate live video set.
And for the live video which is not passed through the auditing, if illegal information is detected to be contained in the live video, immediately stopping playing the live video, prompting a live player, and avoiding the propagation of the live video. The live broadcast environment of the recommended live broadcast video is kept healthy through auditing and filtering, and watching users are guided correctly.
In one embodiment, the video recommendation method further comprises: and sending the recommended content to the user side for displaying.
In step S205, the sorted short videos and the audited live broadcast video are fused to serve as a final recommendation result, and in this step, the fused recommendation result is returned to the front end for display, so that the user can watch and share the recommendation result conveniently.
When the recommendation result is returned to the user side for displaying, different recommendation results can be watched through different operations, for example, when a user watches a certain short video, the user can switch watching contents through left-right sliding, for example, the user slides to the left to watch the previous content and slides to the right to watch the next content, and thus the setting is performed, so that the user can repeatedly watch the watched interesting content through left sliding, the storage effect is achieved, and the display sequence of the recommendation contents in the recommendation pool only needs to be set once before displaying. The sliding operation is merely an illustrative example, and the user may switch the viewing content by other operations.
In one embodiment, as the recommendation result is the fusion of the short video and the live video, the switching between the short video and the live video can be realized through a specific operation instruction, so that the user can switch to the viewing type which is interested by the user at any time, for example, when the user views the video in the APP, the user can switch to the playing page of the live video through three-finger downslide, and switch to the playing page of the short video through two-finger downslide, and simultaneously can display the viewed and unviewed content by combining left-side downslide and right-side downslide.
Switching short videos and live videos through multi-finger sliding is only an implementation mode, and is not a limitation on implementation of the embodiment.
When the user watches the recommended content in the APP, the user can also share the watched content, and the sharing operation can be realized through various ways, such as forwarding to a group or a friend in the APP, forwarding to social platforms such as a WeChat friend circle and a QQ space, or sharing to friends or groups of other APPs such as WeChat. No matter which platform the user shares, when the user shares, the user shares according to the sequence of the existing recommended content in the recommendation pool, for example, the user watches a short video and shares the short video with a WeChat friend circle, so that when other friends in the WeChat watch the short video, the recommended positions of the short video or the live video before and after the short video are the same as the inside of the APP. For example, a user carries out live broadcast in an APP, hopes to obtain more attention, shares live broadcast videos to other social platforms, and when friends on other social platforms watch the live broadcast videos, the live broadcast videos may enter the APP to be watched in order to increase interaction and watching experience, so that the watching interest of the user is increased, the watching selectivity of the user is increased, and the user experience is improved.
When the user watches the recommended content shared by the user in the APP on the sharing page of the social platform, the switching between the short video and the live video can be realized through the operation instruction, for example, the short video is selected to be watched by sliding leftwards, the live video is selected to be watched by sliding rightwards, the watching selectivity of the user is increased, and the user experience and the good sensitivity of the APP are improved.
In the embodiment, after the candidate video set and the candidate live broadcast set are generated, the video recommendation method respectively sorts, audits and screens the candidate video and the live broadcast video, then displays the candidate video and the live broadcast video to the user, ensures the quality of the recommended content by sorting and auditing the candidate video and the live broadcast video, preferentially recommends the video and the live broadcast with higher comprehensive degree to the user, increases the watching interest of the user, ensures the diversity of the recommended content, and improves the user experience.
In one embodiment, the video recommendation method of this embodiment further includes: and updating the recommended content at regular time.
The recommended content comprises a short video and a live video, the live video is changed at a high frequency, the popularity of the short video is changed at any time, and the recommended content needs to be updated at regular time in combination with the current situation in order to adapt to the watching interest of the user. The operation dynamics of recommended contents are concerned in real time for each recommended content, the attention of the recommended contents is returned regularly, the favorite degree of the recommended contents of a user is analyzed, autonomous learning is continuously carried out, the types which are interested by most users are recorded, the recommendation proportion of short videos or live videos of the types is properly increased, and other types are added during each recommendation, so that the diversity of the recommended contents is guaranteed.
By regularly collecting the sharing results of each time and adding the sharing results into the next recommendation as a certain influence factor, iterative computation is continuously carried out, the recommendation content is updated, the preference of the user is adapted, and the current development trend is met.
Fig. 3 is a partial flowchart according to an exemplary embodiment of step S202 in the video recommendation method shown in fig. 2, where step S202 is to generate a candidate video set and a candidate live broadcast set according to historical sharing data, and the present embodiment mainly describes a process of generating a candidate live broadcast set according to sharing data, and specifically includes step S2021-step S2023.
In step S2021, the live IDs and affiliate IDs of all live videos being played at the current time are queried.
The ID is used for marking and identifying the identity of the user, the user can obtain a specific ID as the identity mark when using different APPs, correspondingly, when the user publishes the content in the APP, each content can be distinguished by the corresponding ID. In this embodiment, when the user logs in the APP, the user has an ID number belonging to the user, and when the user publishes the short video or the live video in the APP, the user also obtains the corresponding video ID.
For example, a registered user logs in a client using a user ID, and views and shares a video, and when sharing, some information related to the identity of the user, such as the user ID (user _ ID), may be added to the video link.
When a user carries out live broadcasting in the APP, a user ID used by the user is regarded as a main broadcast ID, a mark of a live video in which the main broadcast ID is carried out is a live broadcast ID, the main broadcast ID and the live broadcast ID in which the main broadcast ID is carried out form a one-to-one correspondence relation, and the main broadcast ID is added into the live broadcast video to indicate affiliation.
And selecting parts from all live broadcast videos at the current moment to be added into a candidate live broadcast set, so that the live broadcast IDs of all live broadcast videos at the current moment and the anchor broadcast IDs for the live broadcast videos at the current moment are required to be obtained. And then the corresponding live broadcast video can be represented by the live broadcast ID, so that the selection of the subsequent live broadcast video is facilitated.
In step S2022, for each piece of live video being played, the attention of the piece of live video is counted.
According to the embodiment, when the live video is selected, the live video to be recommended needs to be selected according to the attention degree obtained by the live video, so that the attention degree obtained by each piece of live video in the process of playing needs to be counted.
Collecting all live broadcast videos in a certain current small time period, acquiring a live broadcast ID of each live broadcast video, inquiring and counting detailed information such as the number of watching people, the number of praise people, real-time popularity and the like corresponding to the live broadcast video, taking the obtained indexes as scoring standards of the attention, and calculating the attention score corresponding to each live broadcast ID by adopting a certain algorithm, for example, the value after the number of the model indexes is weighted represents the attention score. And establishing a table about the live ID, the anchor ID and the attention degree, and arranging all live videos in descending order according to the attention degree scores.
In step S2023, the portion with higher attention is added to the candidate live broadcast set in the live broadcast video.
In the last step, the attention or live broadcast heat corresponding to each live broadcast video is calculated, the live broadcast video with high attention can be considered to be in line with the interests of most people to a certain extent, is popular with users, and has sharing value, so that the part with higher attention score is selected from the table and added into the candidate live broadcast set. For example, the live video with the top 20 ranked attention scores is picked and added into a candidate live set to wait for recommendation.
In one embodiment, the live videos are screened secondarily, a small amount of live videos are selected to be added into a candidate live broadcast set, then types of the live videos with high attention degree scores are counted, the remaining live videos are screened again, and the live videos of the same type are selected to be added into the candidate live broadcast set.
In one embodiment, short videos are selected according to the type of the live video with higher statistical attention degree score, namely, part of short videos of the same type are selected and added into a corresponding candidate video set; or selecting live broadcast videos according to the short video types with higher attention degree scores, selecting a small number of live broadcast videos of the same type from the live broadcast videos according to the short video types with higher attention degree scores, and adding the selected live broadcast videos into a candidate live broadcast set, so that the richness of recommended contents is improved.
In one embodiment, the step of generating the candidate live broadcast set of the present embodiment further includes: and obtaining friend IDs with similar relations to the anchor IDs.
Friends often have more similarity, and many close friends have many common interests, such as WeChat friends, who are often relatives, classmates, colleagues, or friends with some common interests. When a user watches videos shared by friends in a friend circle, the videos shared by the friends are often interested because of the attention of the friends or the high similarity between the friends and the users. Similarly, in this embodiment, when a user shares a short video or a live video in an APP, a friend of the user in the APP is interested in the shared content, and the user can watch the video. Even, content that may be shared among friends with common preferences is also very similar, e.g., beauty makeups people often share some beauty makeup videos; gamers often share some game videos, and go to the game strategy; gourmets will typically share some gourmets; sports enthusiasts can share some fitness videos and the like, so according to the behaviors of users, the sharing content and the similar data of sharing authors are judged, the sharing authors with the same hobbies can be found, and then whether the current sharing authors share short videos or live videos is inquired.
For example, according to the anchor ID corresponding to the live video with higher attention obtained in the above steps, a friend ID similar to the anchor ID is obtained, whether the friend ID is live or not is judged, and if the friend ID is live for sharing, the live video shared by the friend ID is added to the candidate live broadcast set.
In one embodiment, further comprising: and adding part of the friend ID into the candidate live broadcast set in the live broadcast video.
And acquiring friend relation data according to the process, when live video recommendation is required, firstly searching similar friend IDs according to the anchor IDs and the live video shared by the anchor IDs, and then selecting parts from the live videos in which the friend IDs are live to join in a candidate live set.
In one embodiment, further comprising: and establishing a secondary friend relationship of the anchor ID, and adding the part of the secondary friend ID corresponding to the secondary friend relationship into a candidate live broadcast set in the live broadcast video.
The friend ID with similarity to the live ID is a first-level friend relation ID, and the first-level friend ID also has friend relation in the application, and the friends are second-level friends of the anchor ID. For example, if A and B are first-level friends (having an interest in each other), B and C are first-level friends (having an interest in each other), then A and C are second-level friend relationships, multi-level friend relationships, and so on. A part of live videos which are played by the multi-level friend IDs are added into a candidate live set, so that the content of the live videos can be enriched, video recommendation is performed by utilizing the multi-level friend relationships, the richness and diversity of recommended contents can be increased, and a user can have more watching choices.
In one embodiment, the newly added live video is checked, the recalled live video is checked, the newly added similar live video also needs to be checked, the live video which does not pass the check is removed from the candidate live video set, and the rest live video is taken as the recommended live video, so that the richness and the reasonability of the recommended live video are ensured.
Fig. 4 is a flowchart according to an exemplary embodiment of step S205 in the video recommendation method shown in fig. 2, which is a detailed description of a process of merging a candidate video set and a candidate live broadcast set, and mainly includes steps S2051 to S2053.
In step S2051, a video list of all candidate videos after sorting is obtained.
In step S203, the candidate videos in the candidate video set are checked and sorted, where the sorting order is the playing order of the short videos during recommendation, and the sorting method is mentioned in step S203, and is not repeated here, and the sorted candidate videos are arranged in a table to generate a video list. Before fusion, a video list is called first to obtain a plurality of short videos to be recommended, wherein the short videos are videos which are subjected to fine screening and have quality assurance.
In step S2052, the live videos are sorted to obtain a live list.
In step S204, the live video in progress in the candidate live broadcast set is audited and filtered, live video that fails in auditing is excluded, and live video with higher attention is retained. In the step, videos in the candidate live broadcast set are sorted according to the historical interest and live broadcast real-time popularity of the user, if the basis in sorting is the same as the judgment standard of the attention, the videos are directly sorted according to the attention scores, if the basis in sorting is different from the judgment standard of the attention, the videos are reordered to form a recommendation sequence of the live broadcast videos, a live broadcast list is obtained after sorting, and personalized live broadcast videos are recommended to the user according to the sorted live broadcast list result.
In step S2053, the live list is merged into the video list and recommended as recommended content.
Inserting live videos in a live list into a video list to be recommended, inserting the live videos according to the sequence of the live list when inserting the live videos, and inserting one or more live videos into adjacent candidate videos in an interval insertion mode or inserting one or more live videos into the candidate videos with the same number every interval; or inserting the candidate videos in the video list to be recommended in the same method in the video list.
When the live list and the video list are fused, other insertion modes can be adopted, for example, the live list and the video list are divided according to types, and the short videos of the same type and the live videos of the same type are put together, and the mode needs to be comprehensively considered by combining with the current sequencing order.
When the fusion is carried out, the number of the candidate videos and the number of the live videos are not required, and the candidate videos and the live videos are adjusted according to actual requirements.
When video recommendation is carried out, recommendation of live video with high popularity is added, the interestingness of the recommended video can be increased, more users are attracted, the users can know news dynamics in real time, and watching experience is improved.
Fig. 5 is a schematic diagram illustrating a video recommendation apparatus according to an example embodiment. The video recommendation apparatus 500 includes an obtaining module 501, a candidate module 502, a ranking module 503, and a recommendation module 504.
The obtaining module 501 is configured to obtain historical sharing data of a user;
the candidate module 502 is configured to generate a candidate video set and a candidate live broadcast set according to the historical sharing data;
the sorting module 503 is configured to sort all candidate videos in the candidate video set;
the recommending module 504 is configured to fuse the sorted candidate video set and the candidate live broadcast set to recommend content as recommended content.
The video recommendation device of the embodiment generates the candidate video set and the candidate live broadcast set according to the historical sharing data by acquiring the historical sharing data of the user, and then, the candidate video set and the candidate live broadcast set are fused and recommended as recommended content, so that live broadcast content is added in video sharing, the recommendation diversity of the recommended content of a sharing page is increased, the sharing page is added in a live broadcast mode, the user is more willing to enter APP internal activities, and the watching experience and the interest of the user are improved.
In one embodiment, the history sharing data includes history sharing content and attention obtained by the history sharing content, the history sharing content includes a history sharing video and an on-air live video, and the candidate module 502 is configured to select a part of the history sharing video and a part of the on-air live video to join the candidate video set and the candidate live set, respectively.
FIG. 6 is a schematic diagram of an aggregated video recommender shown in accordance with an exemplary embodiment.
Fig. 6 is an optimization of the embodiment of fig. 5, and the video recommendation apparatus 600 further includes, in addition to the obtaining module 501, the candidate module 502, the ranking module 503, and the recommendation module 504: an audit module 601.
The auditing module 601 is configured to audit all live videos in the candidate live broadcast set, and remove live videos that are not approved.
Optionally, after the candidate module 502 selects a part of the history sharing videos to add into the candidate video set, the sorting module 503 sorts the candidate videos; after the candidate module 502 selects part of the live broadcast video to be added into the candidate live broadcast set, the auditing module 601 audits the live broadcast video. And then the recommending module 504 fuses the sorted candidate video set and the checked candidate live broadcast set to recommend the candidate video set and the checked candidate live broadcast set as recommended content.
Optionally, the candidate module 502 includes a candidate live broadcast set generation module (not shown in the figure) and a candidate video set generation module (not shown in the figure), wherein the candidate live broadcast set generation module includes:
the query module is used for querying the live broadcast IDs and the affiliated anchor IDs of all live broadcast videos at the current moment;
the statistic module is used for counting the attention of each piece of live broadcast video; and
and the first selection module is used for adding the part with higher attention into the candidate live broadcast set in the live broadcast video.
In one embodiment, the candidate live set generation module of the candidate module 502 further comprises:
the friend relation obtaining module is used for obtaining friend IDs with similar relations to the anchor IDs; and
and the second selection module is used for adding the part of the friend ID into the candidate live broadcast set in the live broadcast video.
In one embodiment, the recommendation module 504 includes:
a video list obtaining module (not shown in the figure) for obtaining a video list of all the candidate videos after sorting;
a live broadcast list acquisition module (not shown in the figure) for sequencing the live broadcast video to obtain a live broadcast list;
and a fusion module (not shown in the figure) for fusing the live list into the video list to be recommended as the recommended content.
In the embodiment, after the candidate video set and the candidate live broadcast set are generated, the video recommendation device respectively sorts, audits and screens the candidate video and the live broadcast video, displays the candidate video and the live broadcast video to the user, ensures the quality of the recommended content by sorting and auditing the candidate video and the live broadcast video, preferentially recommends the video and the live broadcast with higher comprehensive degree to the user, increases the watching interest of the user, ensures the diversity of the recommended content, and improves the user experience.
With regard to the video recommendation apparatus in the above embodiment, since the functions of the respective modules have been described in detail in the above embodiment of the video recommendation method, a relatively brief description is made.
FIG. 7 is a schematic diagram of an aggregated video recommender shown in accordance with an exemplary embodiment.
The video recommendation apparatus 700 of the present embodiment includes an offline module 710 and a server 720. The offline module 710 includes a log storage module 711, an offline data storage module 712, a sample module 713, and a training module 714; the server 720 includes a ranking module 721, a screening module 722, a fusion module 723, a sharing module 724, and a front-end presentation module 725.
The log storage module 711 is configured to collect shared video and live broadcast information, record all operations (clicking, watching, commenting, collecting, forwarding, and the like) of the user on the short video and live broadcast in the service layer, and store the records in a log form, for example, a user click behavior log;
the offline data storage module 712 is configured to collect candidate videos according to the shared video content, collect all live videos in the broadcast according to the stored log information, and add the candidate videos and the live videos in the broadcast into a video library as a preliminary recall result;
the sample module 713 is used for selecting part of videos from the collected videos to be added into the candidate video set according to the records of all operations of the short videos and the live broadcasts of the user on the service layer, and selecting part of videos which are currently live broadcasts to be added into the candidate live broadcast set;
the training module 714 is used for training by adopting a large amount of data, and autonomously learning video and live broadcast selection;
after selection, a selection model is generated for selecting videos from different video sharing pages and adding live broadcasts into a candidate set, and then the server 720 adopts the model to perform fine screening on the candidate video set and the candidate live broadcast set collected by the model.
The sorting module 721 is used for extracting the behavior characteristics of the user and sorting the videos and live broadcasts in the candidate set;
the screening module 722 is configured to perform auditing and filtering on live videos in the candidate live broadcast set, and delete live broadcasts that cannot be audited according to a live broadcast auditing result;
the fusion module 723 is configured to fuse the video sorting result and the live broadcast audit result as recommended content;
the sharing module 724 is mainly responsible for data communication with the service layer, receiving a data request and sending a recommendation result, for example, sharing recommended content to different pages;
the front-end presentation module 725 returns the merged recommended content to the front end for presentation to the user.
Fig. 8 is a block diagram illustrating an electronic device 1200 for use in the video recommendation method described above, according to an example embodiment. For example, the electronic device 1200 may be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, an exercise device, a personal digital assistant, and the like.
Referring to fig. 8, electronic device 1200 may include one or more of the following components: a processing component 1202, a memory 1204, a power component 1206, a multimedia component 1208, an audio component 1210, an input/output (I/O) interface 1212, a sensor component 1214, and a communications component 1216.
The processing component 1202 generally controls overall operation of the electronic device 1200, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing components 1202 may include one or more processors 1220 to execute instructions to perform all or a portion of the steps of the methods described above. Further, the processing component 1202 can include one or more modules that facilitate interaction between the processing component 1202 and other components. For example, the processing component 1202 can include a multimedia module to facilitate interaction between the multimedia component 1208 and the processing component 1202.
The memory 1204 is configured to store various types of data to support operation at the electronic device 1200. Examples of such data include instructions for any application or method operating on the electronic device 1200, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 1204 may be implemented by any type or combination of volatile or non-volatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.
The power supply component 1206 provides power to the various components of the electronic device 1200. The power components 1206 may include a power management system, one or more power sources, and other components associated with generating, managing, and distributing power for the electronic device 1200.
The multimedia component 1208 comprises a screen providing an output interface between the electronic device 1200 and a user. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 1208 includes a front facing camera and/or a rear facing camera. The front camera and/or the rear camera may receive external multimedia data when the electronic device 1200 is in an operation mode, such as a shooting mode or a video mode. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.
Audio component 1210 is configured to output and/or input audio signals. For example, the audio assembly 1210 includes a Microphone (MIC) configured to receive external audio signals when the electronic device 1200 is in an operational mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals may further be stored in the memory 1204 or transmitted via the communication component 1216. In some embodiments, audio assembly 1210 further includes a speaker for outputting audio signals.
The I/O interface 1212 provides an interface between the processing component 1202 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a start button, and a lock button.
The sensor assembly 1214 includes one or more sensors for providing various aspects of state assessment for the electronic device 1200. For example, the sensor assembly 1214 may detect an open/closed state of the electronic device 1200, the relative positioning of components, such as a display and keypad of the apparatus 1200, the sensor assembly 1214 may also detect a change in the position of the electronic device 1200, or a component of the electronic device 1200, the presence or absence of user contact with the electronic device 1200, the orientation or acceleration/deceleration of the electronic device 1200, and a change in the temperature of the electronic device 1200. The sensor assembly 1214 may include a proximity sensor configured to detect the presence of a nearby object in the absence of any physical contact. The sensor assembly 1214 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 1214 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
The communications component 1216 is configured to facilitate communications between the electronic device 1200 and other devices in a wired or wireless manner. The electronic device 1200 may access a wireless network based on a communication standard, such as WiFi, a carrier network (such as 2G, 3G, 4G, or 5G), or a combination thereof. In an exemplary embodiment, the communication component 1216 receives the broadcast signal or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communications component 1216 further includes a Near Field Communication (NFC) module to facilitate short-range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.
In an exemplary embodiment, the electronic device 1200 may be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors or other electronic components for performing the above-described video recommendation method.
In an exemplary embodiment, a non-transitory computer readable storage medium comprising instructions, such as the memory 1204 comprising instructions, executable by the processor 1220 of the electronic device 1200 to perform the above-described method is also provided. For example, the non-transitory computer readable storage medium may be a ROM, a Random Access Memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.
Fig. 9 is a block diagram illustrating a video recommendation apparatus 1300 for the above-described video recommendation method according to an exemplary embodiment. For example, the apparatus 1300 may be provided as a server. Referring to fig. 9, apparatus 1300 includes a processing component 1322, which further includes one or more processors, and memory resources, represented by memory 1332, for storing instructions, such as application programs, that may be executed by processing component 1322. The application programs stored in memory 1332 may include one or more modules that each correspond to a set of instructions. Further, processing component 1322 is configured to execute instructions to perform the video recommendation method described above.
The apparatus 1300 may also include a power component 1326 configured to perform power management for the apparatus 1300, a wired or wireless network interface 1350 configured to connect the apparatus 1300 to a network, and an input-output (I/O) interface 1358. The apparatus 1300 may operate based on an operating system stored in the memory 1332, such as Windows Server, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, or the like.
Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the application being indicated by the following claims.
It will be understood that the present application is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the application is limited only by the appended claims.
Claims (18)
1. A method for video recommendation, comprising:
acquiring historical sharing data of a user;
generating a candidate video set and a candidate live broadcast set according to the historical sharing data;
ranking all candidate videos in the candidate video set; and
fusing the sorted candidate video set and the candidate live broadcast set to be recommended as recommended content,
the candidate live set includes a plurality of on-air live videos shared by an anchor.
2. The video recommendation method according to claim 1, wherein the history sharing data includes history sharing content and attention obtained by the history sharing content.
3. The video recommendation method according to claim 2, wherein the history shared content comprises a history shared video and an on-air live video, and a part of the history shared video and a part of the on-air live video are respectively added to the candidate video set and the candidate live video set according to a preset rule.
4. The video recommendation method of claim 1, further comprising: and auditing all on-broadcast live videos in the candidate live broadcast set, and removing the on-broadcast live videos which are not approved.
5. The video recommendation method according to claim 4, wherein an audit time is set, an audit state of each live video is monitored, and if an instruction that the audit is passed is not received within the audit time, the live video is deleted from the candidate live video set.
6. The video recommendation method of claim 1, wherein generating a candidate live set from the historical sharing data comprises:
inquiring live broadcast IDs and affiliated anchor IDs of all live broadcast videos at the current moment;
counting attention of each live broadcast video; and
and adding the part of the on-air live video with higher attention into the candidate live video set.
7. The video recommendation method of claim 6, further comprising:
obtaining friend IDs with similar relations to the anchor IDs;
and adding part of the friend ID into the candidate live broadcast set in the live broadcast video.
8. The video recommendation method according to claim 4, wherein the step of recommending the candidate video set and the candidate live broadcast set after being fused as recommended content comprises:
acquiring a video list of all the candidate videos after sequencing;
sequencing the live broadcast videos to obtain a live broadcast list;
and the live broadcast list is merged into the video list to be used as the recommended content for recommendation.
9. A video recommendation apparatus, comprising:
the acquisition module is used for acquiring historical shared data of a user;
the candidate module is used for generating a candidate video set and a candidate live broadcast set according to the historical sharing data;
the sorting module is used for sorting all the candidate videos in the candidate video set; and
a recommending module for fusing the sorted candidate video set and the candidate live broadcast set to recommend the content as recommended content,
the candidate live set includes a plurality of on-air live videos shared by an anchor.
10. The video recommendation device according to claim 9, wherein the history sharing data includes history sharing content and attention obtained by the history sharing content.
11. The video recommendation device according to claim 10, wherein the history shared content includes a history shared video and an on-air video, and a part of the history shared video and a part of the on-air video are selected and added to the candidate video set and the candidate on-air video set respectively.
12. The video recommendation device of claim 9, further comprising: and the auditing module is used for auditing all live broadcast videos in the candidate live broadcast set and removing the live broadcast videos which are not approved.
13. The video recommendation device of claim 12, wherein an audit time is set, an audit status of each live video is monitored, and if no instruction that the audit is passed is received within the audit time, the live video is deleted from the candidate live video set.
14. The video recommendation device of claim 9, wherein the candidate modules comprise:
the query module is used for querying the live broadcast IDs and the affiliated anchor IDs of all live broadcast videos at the current moment;
the statistic module is used for counting the attention of each piece of live broadcast video;
and the first selection module is used for adding the part of the on-air live video with higher attention into the candidate live video set.
15. The video recommendation device of claim 14, wherein the candidate modules further comprise:
a friend relation obtaining module, configured to obtain a friend ID having a similar relation to the anchor ID; and
and the second selection module is used for adding part of the friend ID into the candidate live broadcast set in the live broadcast video.
16. The video recommendation device of claim 12, wherein the recommendation module comprises:
the video list acquisition module is used for acquiring a video list of all the candidate videos after sequencing;
the live broadcast list acquisition module is used for sequencing the live broadcast video to obtain a live broadcast list;
and the fusion module is used for fusing the live list into the video list to serve as the recommended content for recommendation.
17. An electronic device, comprising:
a processor;
a memory for storing processor-executable instructions;
wherein the processor is configured to perform the video recommendation method of any of the preceding claims 1-8.
18. A computer-readable storage medium storing computer instructions which, when executed by a processor, implement the video recommendation method of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811588951.XA CN109769128B (en) | 2018-12-25 | 2018-12-25 | Video recommendation method, video recommendation device and computer-readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811588951.XA CN109769128B (en) | 2018-12-25 | 2018-12-25 | Video recommendation method, video recommendation device and computer-readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109769128A CN109769128A (en) | 2019-05-17 |
| CN109769128B true CN109769128B (en) | 2021-07-13 |
Family
ID=66451571
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811588951.XA Active CN109769128B (en) | 2018-12-25 | 2018-12-25 | Video recommendation method, video recommendation device and computer-readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109769128B (en) |
Families Citing this family (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110297942B (en) * | 2019-06-26 | 2024-01-05 | 广州市百果园信息技术有限公司 | Video exploration method, device, equipment and storage medium |
| CN112218114B (en) * | 2019-07-10 | 2023-01-03 | 中国电信股份有限公司 | Video cache control method, device and computer readable storage medium |
| CN110366043B (en) * | 2019-08-20 | 2022-02-18 | 北京字节跳动网络技术有限公司 | Video processing method and device, electronic equipment and readable medium |
| CN112565797B (en) * | 2019-09-10 | 2022-08-05 | 北京达佳互联信息技术有限公司 | Page request processing method and device, electronic equipment and medium |
| CN110489655A (en) * | 2019-09-16 | 2019-11-22 | 浙江同花顺智能科技有限公司 | Hot content determination, recommended method, device, equipment and readable storage medium storing program for executing |
| CN112565814B (en) * | 2019-09-26 | 2022-06-14 | 腾讯科技（深圳）有限公司 | Video distribution method, device, storage medium and server |
| CN112749324B (en) * | 2019-10-31 | 2024-03-12 | 北京达佳互联信息技术有限公司 | Network medium information delivery control method and device, electronic equipment and storage medium |
| CN110866805A (en) * | 2019-11-13 | 2020-03-06 | 第四范式（北京）技术有限公司 | Method and system for recommending object |
| CN110856031B (en) * | 2019-11-18 | 2022-01-18 | 广州市百果园信息技术有限公司 | Media resource display system, method, equipment and storage medium |
| CN111212299B (en) * | 2020-01-16 | 2022-02-11 | 广州酷狗计算机科技有限公司 | Method and device for acquiring live video tutorial, server and storage medium |
| CN111523050B (en) * | 2020-04-16 | 2023-09-19 | 咪咕文化科技有限公司 | Content recommendation method, server and storage medium |
| CN111737566A (en) * | 2020-05-15 | 2020-10-02 | 深圳市战音科技有限公司 | Work sorting method, device and storage medium |
| CN111614998B (en) * | 2020-05-26 | 2022-03-18 | 上海连尚网络科技有限公司 | Method for processing promotion video, electronic device and computer-readable storage medium |
| CN112051949A (en) * | 2020-09-14 | 2020-12-08 | 北京达佳互联信息技术有限公司 | Content sharing method and device and electronic equipment |
| CN112637685B (en) * | 2020-12-11 | 2024-01-30 | 上海连尚网络科技有限公司 | Video processing method and device |
| CN112714327B (en) * | 2020-12-21 | 2022-02-25 | 腾讯科技（深圳）有限公司 | Interaction method, device and equipment based on live application program and storage medium |
| CN112650884B (en) * | 2020-12-21 | 2022-04-01 | 四川长虹电器股份有限公司 | Video pushing method based on deep learning |
| CN112765478B (en) * | 2021-03-12 | 2023-06-20 | 北京百度网讯科技有限公司 | Method, apparatus, device, medium and program product for recommending content |
| CN113205382B (en) * | 2021-04-30 | 2024-02-20 | 北京有竹居网络技术有限公司 | Method, apparatus, device, storage medium and program product for determining object |
| CN113794910B (en) * | 2021-09-14 | 2024-04-19 | 广州方硅信息技术有限公司 | Video pushing method, device, electronic equipment and storage medium |
| CN113886633A (en) * | 2021-10-11 | 2022-01-04 | 平安国际智慧城市科技股份有限公司 | Video recommendation, device, equipment and storage medium based on artificial intelligence |
| CN114430503B (en) * | 2022-01-25 | 2023-08-04 | 上海影宴数码科技有限公司 | Short video-based big data superposition recommendation method |
| CN114520929A (en) * | 2022-02-13 | 2022-05-20 | 武汉斗鱼鱼乐网络科技有限公司 | Video playing method, device, medium and computer equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070124493A1 (en) * | 2005-10-07 | 2007-05-31 | The Board Of Trustees Of The Leland Stanford Jr. University | System and Method for Ordering Content for User Review |
| CN102843602A (en) * | 2011-06-22 | 2012-12-26 | 宏碁股份有限公司 | Television system and method for providing program recommendation list of television system |
| CN107436914A (en) * | 2017-06-06 | 2017-12-05 | 北京小度信息科技有限公司 | Recommend method and device |
| CN108388629A (en) * | 2018-02-13 | 2018-08-10 | 北京奇虎科技有限公司 | A kind of multimedia information push method, device and electronic equipment |
| CN108419088A (en) * | 2018-02-08 | 2018-08-17 | 华南理工大学 | A kind of recommendation of the channels method towards high sudden user's request |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105224576A (en) * | 2014-07-01 | 2016-01-06 | 上海视畅信息科技有限公司 | A kind of video display intelligent recommendation method |
-
2018
- 2018-12-25 CN CN201811588951.XA patent/CN109769128B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070124493A1 (en) * | 2005-10-07 | 2007-05-31 | The Board Of Trustees Of The Leland Stanford Jr. University | System and Method for Ordering Content for User Review |
| CN102843602A (en) * | 2011-06-22 | 2012-12-26 | 宏碁股份有限公司 | Television system and method for providing program recommendation list of television system |
| CN107436914A (en) * | 2017-06-06 | 2017-12-05 | 北京小度信息科技有限公司 | Recommend method and device |
| CN108419088A (en) * | 2018-02-08 | 2018-08-17 | 华南理工大学 | A kind of recommendation of the channels method towards high sudden user's request |
| CN108388629A (en) * | 2018-02-13 | 2018-08-10 | 北京奇虎科技有限公司 | A kind of multimedia information push method, device and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109769128A (en) | 2019-05-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109769128B (en) | Video recommendation method, video recommendation device and computer-readable storage medium | |
| CN111556327A (en) | Live broadcast room recommendation method, device, terminal, server, system and storage medium | |
| US8578415B2 (en) | Information providing method, content display terminal, portable terminal, server device, information providing system and program | |
| CN106874308B (en) | Recommendation method and device and recommendation device | |
| CN107209776B (en) | Methods, systems, and media for presenting information related to an event based on metadata | |
| CN111258435B (en) | Comment method and device for multimedia resources, electronic equipment and storage medium | |
| CN110800012A (en) | Generation of interactive content with advertisements | |
| CN109688479B (en) | Bullet screen display method, bullet screen display device and bullet screen display server | |
| CN103718166A (en) | Information processing apparatus, information processing method, and computer program product | |
| CN111859020B (en) | Recommendation method, recommendation device, electronic equipment and computer readable storage medium | |
| US20170118298A1 (en) | Method, device, and computer-readable medium for pushing information | |
| CN111556366A (en) | Multimedia resource display method, device, terminal, server and system | |
| CN111783001A (en) | Page display method and device, electronic equipment and storage medium | |
| JP6988961B2 (en) | Information processing system and information processing method | |
| CN109783656B (en) | Recommendation method and system of audio and video data, server and storage medium | |
| CN110688576B (en) | Content recommendation method and device, electronic equipment and storage medium | |
| CN106649781B (en) | Application recommendation method and device | |
| CN112464031A (en) | Interaction method, interaction device, electronic equipment and storage medium | |
| US20140195647A1 (en) | Bookmarking system | |
| US11256382B2 (en) | Method for displaying interactive content, electronic device, and storage medium | |
| JP2011223571A (en) | Information processing unit, information processing system and program | |
| CN112131466A (en) | Group display method, device, system and storage medium | |
| CN111046268B (en) | Information recommendation method and device and electronic equipment | |
| CN112685641B (en) | Information processing method and device | |
| US20200228617A1 (en) | Event recording method and apparatus, electronic device, 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |