CN112487240A - Video data recommendation method and device - Google Patents
Video data recommendation method and device Download PDFInfo
- Publication number
- CN112487240A CN112487240A CN202011206237.7A CN202011206237A CN112487240A CN 112487240 A CN112487240 A CN 112487240A CN 202011206237 A CN202011206237 A CN 202011206237A CN 112487240 A CN112487240 A CN 112487240A
- Authority
- CN
- China
- Prior art keywords
- video
- demand
- recommendation
- sequence data
- target user
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000012549 training Methods 0.000 claims abstract description 97
- 238000012544 monitoring process Methods 0.000 claims abstract description 10
- 238000007781 pre-processing Methods 0.000 claims description 29
- 238000004590 computer program Methods 0.000 claims description 8
- 230000001965 increasing effect Effects 0.000 claims description 8
- 230000005540 biological transmission Effects 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 6
- 230000002860 competitive effect Effects 0.000 description 5
- 238000012552 review Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000012217 deletion Methods 0.000 description 2
- 230000037430 deletion Effects 0.000 description 2
- 230000002708 enhancing effect Effects 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 230000003321 amplification Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000036961 partial effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000003442 weekly effect Effects 0.000 description 1
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
- G06F16/70—Information retrieval; Database structures therefor; File system structures therefor of video data
- G06F16/73—Querying
- G06F16/735—Filtering based on additional data, e.g. user or group profiles
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The embodiment of the application provides a recommendation method and device for video data, wherein the method comprises the following steps: monitoring a video data recommendation request containing identification information of a target user; acquiring video sequence data of a target user according to the identification information; inputting the video sequence data into a video recommendation model and outputting a video recommendation result; obtaining at least one corresponding second video according to the similarity of each first video in the video recommendation result; and inserting the second video into the video recommendation result to obtain a new video recommendation result. The embodiment of the application does not depend on the labels, the dimensions and the like of the videos, does not need a large number of videos, is suitable for the videos without the conditions of the labels, the dimensions and the like, such as training videos and the like, and improves the recommendation accuracy rate of the videos such as the training videos. The new video with high similarity can be inserted into the video recommendation result, so that the purposes of recommending the old video and the new video are achieved, and the recommendation rate of the new video is improved.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a method and an apparatus for recommending video data.
Background
With the development of deep learning and big data technology, video recommendation schemes are more and more intelligent, but generally, recommendation models with better effects need to rely on more video data and more video tags, and accordingly, more privacy of customers needs to be acquired, and higher cost is spent.
In many video recommendation scenes (such as training video recommendation), the dimensionality of video data is not high, the video data does not have tags, the data volume of the video data is relatively small, and if a traditional video recommendation scheme is used for recommending training videos, the problem of low recommendation accuracy is caused.
Disclosure of Invention
In view of the above problems, embodiments of the present application are proposed to provide a recommendation method and apparatus for video data that overcome or at least partially solve the above problems.
In order to solve the above problem, according to a first aspect of an embodiment of the present application, a method for recommending video data is disclosed, including: monitoring a video data recommendation request, wherein the video data recommendation request contains identification information of a target user; acquiring video sequence data of the target user according to the identification information, wherein the video sequence data is generated based on a plurality of video-on-demand records which are sequentially arranged; inputting the video sequence data into a video recommendation model, and outputting a video recommendation result, wherein the video recommendation result comprises a plurality of first videos which are sequentially arranged; acquiring at least one corresponding second video according to the similarity of each first video, wherein the similarity between the second video and the corresponding first video is greater than a preset similarity threshold, and the earliest on-demand time information of the second video is positioned behind the earliest on-demand time information of the corresponding first video; and inserting the second video into the video recommendation result to obtain a new video recommendation result.
Optionally, the inserting the second video into the video recommendation result to obtain a new video recommendation result includes: inserting the second video after the position of the corresponding first video in the video recommendation result, and deleting at least one first video which is reciprocal in the video recommendation result; wherein the number of the deleted first video data is the same as the number of the second video data inserted into the video recommendation result.
Optionally, the method further comprises: acquiring the on-demand states of the first video and the second video; displaying corresponding on-demand states at preset positions of the first video and the second video, and classifying the first video and the second video according to the on-demand states.
Optionally, the acquiring the video sequence data of the target user according to the identification information includes: searching to obtain the video-on-demand record of the target user according to the identification information; carrying out preprocessing operation on the video-on-demand record of the target user; generating the video sequence data according to the video-on-demand record of the target user after the preprocessing operation; or, according to the identification information, the video-on-demand record of the target user is obtained without searching; and taking preset default video sequence data as the video sequence data of the target user.
Optionally, the default video sequence data is generated by: acquiring video on demand records of all users; performing the preprocessing operation on the video-on-demand records of all the users; generating all video sequence data of all users according to the video on demand records of all users after the preprocessing operation; and taking the video sequence data with the highest frequency in the video sequence data of all the users as the default video sequence data.
Optionally, the method further comprises: counting the on-demand times of each video in the video on-demand records of all users; increasing the video-on-demand times of the videos of which the video-on-demand times are smaller than a preset time threshold value to obtain a new video-on-demand record; and training according to the new video on demand record to obtain the video recommendation model.
Optionally, the acquiring the video-on-demand records of all the users includes: and acquiring the latest multiple video-on-demand records of all the users.
According to a second aspect of the embodiments of the present application, there is also disclosed a recommendation apparatus for video data, including: the monitoring module is configured to monitor a video data recommendation request, and the video data recommendation request contains identification information of a target user; an acquisition module configured to acquire video sequence data of the target user according to the identification information, the video sequence data being generated based on a plurality of pieces of video-on-demand records arranged in sequence; the transmission module is configured to input the video sequence data into a video recommendation model and output a video recommendation result, wherein the video recommendation result comprises a plurality of first videos which are sequentially arranged; the obtaining module is further configured to obtain at least one corresponding second video according to the similarity of each first video, the similarity between the second video and the corresponding first video is greater than a preset similarity threshold, and the earliest on-demand time information of the second video is located after the earliest on-demand time information of the corresponding first video; and the updating module is configured to insert the second video into the video recommendation result to obtain a new video recommendation result.
Optionally, the update module is configured to insert the second video after the position of the corresponding first video in the video recommendation, and delete at least one of the first videos in the video recommendation that is reciprocal; wherein the number of the deleted first video data is the same as the number of the second video data inserted into the video recommendation result.
Optionally, the obtaining module is further configured to obtain on-demand statuses of the first video and the second video; the device further comprises: a presentation module configured to present corresponding on-demand states at preset positions of the first video and the second video, and classify the first video and the second video according to the on-demand states.
Optionally, the obtaining module includes: the searching module is configured to search for the video-on-demand record of the target user according to the identification information; a preprocessing module configured to perform a preprocessing operation on the video-on-demand record of the target user; a generation module configured to generate the video sequence data according to the video-on-demand record of the target user after the preprocessing operation; or, the searching module is configured to obtain the video-on-demand record of the target user according to the non-searched identification information; a determination module configured to use preset default video sequence data as the video sequence data of the target user.
Optionally, the obtaining module is further configured to obtain video-on-demand records of all users; the preprocessing module is further configured to perform the preprocessing operation on the video-on-demand records of all the users; the generation module is further configured to generate each video sequence data of all the users according to the video-on-demand records of all the users after the preprocessing operation; the determining module is further configured to use, as the default video sequence data, video sequence data that appears most frequently among the video sequence data of all the users.
Optionally, the apparatus further comprises: the counting module is configured to count the on-demand times of each video in the video on-demand records of all users; the increasing module is configured to increase the video-on-demand times of videos of which the video-on-demand times are smaller than a preset time threshold value to obtain a new video-on-demand record; and the training module is configured to obtain the video recommendation model according to the new video-on-demand record training.
Optionally, the obtaining module is configured to obtain the latest multiple pieces of vod records of all the users.
The embodiment of the application has the following advantages:
the embodiment of the application provides a recommendation scheme of video data, wherein a video data recommendation request containing identification information of a target user is monitored, video sequence data of the target user are obtained according to the identification information of the target user, and the video sequence data are generated by a plurality of pieces of video-on-demand records which are sequentially arranged. Then, the video sequence data is input to a video recommendation model, and a video recommendation result including a plurality of first videos arranged in sequence is output. Further, a corresponding second video is obtained according to the similarity of each first video, wherein the similarity between the second video and the corresponding first video is greater than a preset similarity threshold, and the earliest on-demand time information of the second video is located behind the earliest on-demand time information of the corresponding first video. After the second video is obtained, the second video is inserted into the video recommendation result, and a new video recommendation result is obtained.
In the embodiment of the application, on one hand, a video recommendation result is output according to video sequence data generated by video on demand recording and a video recommendation model. In the process of video recommendation, the method does not depend on the labels, the dimensions and the like of the videos, does not need a large number of videos, is suitable for the videos without the conditions of the labels, the dimensions and the like, such as training videos and the like, and improves the recommendation accuracy of the videos such as the training videos. On the other hand, a new video with high similarity, namely the second video, can be inserted into the video recommendation result, so that the purposes of recommending the old video, namely the first video, and recommending the new video are achieved, and the recommendation rate of the new video is improved.
Drawings
FIG. 1 is a flowchart illustrating steps of an embodiment of a method for recommending video data according to the present application;
FIG. 2 is a flowchart illustrating steps of an embodiment of a method for recommending training videos according to the present application;
FIG. 3 is a schematic flow chart diagram illustrating an embodiment of a training video recommendation method of the present application;
fig. 4 is a block diagram illustrating a configuration of an embodiment of a video data recommendation apparatus according to the present application;
FIG. 5 is a schematic block diagram of a computer system suitable for use in implementing an electronic device according to embodiments of the present application.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, the present application is described in further detail with reference to the accompanying drawings and the detailed description.
Referring to fig. 1, a flowchart illustrating steps of an embodiment of a method for recommending video data according to the present application is shown. The video data recommendation method can be applied to a server, and specifically can include the following steps:
In an embodiment of the present application, the video data recommendation request may come from a terminal communicatively connected to the server. In practical application, a target user can log in a terminal and access a video recommendation page, the terminal can respond to the access operation of the target user to obtain identification information of the target user, the identification information is used for uniquely representing the target user, the identification information can be a user name, a number and the like, and the content, the format and the like of the identification information are not specifically limited in the embodiment of the application. The terminal may generate a video data recommendation request according to the identification information of the target user, where the video data recommendation request may carry the identification information of the target user, and may also carry address information of the terminal, and the like. The content, format and the like of the video data recommendation request are not particularly limited in the embodiments of the present application.
And 102, acquiring video sequence data of the target user according to the identification information.
In embodiments of the present application, the video sequence data may be generated based on a plurality of video-on-demand recordings arranged in a sequence. Each video on demand record may include, but is not limited to: on-demand start time, on-demand end time, video name, video number, video on-demand duration, total video duration, etc. In practical application, the video sequence data can be searched according to the identification information of the target user.
And 103, inputting the video sequence data into a video recommendation model, and outputting a video recommendation result.
In the embodiment of the application, the video recommendation model can be configured in the server, and the video recommendation model can be periodically trained in the server so as to improve the recommendation accuracy of the video recommendation model. For example, the video recommendation model may be trained weekly or daily. The output video recommendation result may include a plurality of first videos arranged in sequence. In practical application, the first video with the front position in the video recommendation result is more in line with the conditions of the target user such as the on-demand interest and the on-demand habit relative to the first video with the back position.
And 104, acquiring at least one corresponding second video according to the similarity of each first video.
In an embodiment of the present application, the similarity indicates a similarity between a first video and a second video, and the obtained similarity between the second video and a corresponding first video should be greater than a preset similarity threshold. For example, the first video is v01, the second video is v02, the similarity between the second video v02 and the first video v01 is 1, and the preset similarity threshold is 0.9. The second video v02 may be the second video corresponding to the first video v 01. Furthermore, the earliest on-demand time information of the second video v02 needs to be located after the earliest on-demand time information of the first video v 01. If the earliest on-demand time information of the second video v02 is 10, 01, 10 in 2020: 00, the earliest on-demand time information of the first video v01 is 09 month, 20 day 10 in 2020: 00. that is, the second video is a new video relative to the first video, which is an old video relative to the second video.
And 105, inserting the second video into the video recommendation result to obtain a new video recommendation result.
In the embodiment of the present application, as described above, the second video v02 is inserted into the video recommendation result, specifically, the second video v02 may be inserted into a position behind the first video v01, so as to obtain a new video recommendation result. The server can return the new video recommendation result to the terminal so as to display the new video recommendation result on the terminal. The target user can play any first video or second video in the new video recommendation result on the terminal.
The embodiment of the application provides a recommendation scheme of video data, wherein a video data recommendation request containing identification information of a target user is monitored, video sequence data of the target user are obtained according to the identification information of the target user, and the video sequence data are generated by a plurality of pieces of video-on-demand records which are sequentially arranged. Then, the video sequence data is input to a video recommendation model, and a video recommendation result including a plurality of first videos arranged in sequence is output. Further, a corresponding second video is obtained according to the similarity of each first video, wherein the similarity between the second video and the corresponding first video is greater than a preset similarity threshold, and the earliest on-demand time information of the second video is located behind the earliest on-demand time information of the corresponding first video. After the second video is obtained, the second video is inserted into the video recommendation result, and a new video recommendation result is obtained.
In the embodiment of the application, on one hand, a video recommendation result is output according to video sequence data generated by video on demand recording and a video recommendation model. In the process of video recommendation, the method does not depend on the labels, the dimensions and the like of the videos, does not need a large number of videos, is suitable for the videos without the conditions of the labels, the dimensions and the like, such as training videos and the like, and improves the recommendation accuracy of the videos such as the training videos. On the other hand, a new video with high similarity, namely the second video, can be inserted into the video recommendation result, so that the purposes of recommending the old video, namely the first video, and recommending the new video are achieved, and the recommendation rate of the new video is improved.
In an exemplary embodiment of the present application, in the process of executing step 105, the second video may be inserted after the position of the corresponding first video in the video recommendation, and at least one first video that is the reciprocal in the video recommendation is deleted; wherein the number of the deleted first video data is the same as the number of the second video data inserted into the video recommendation result. For example, the video recommendation result includes 20 first videos, which are v1, v2, v3, … … and v20 in sequence. The similarity of the second video vv2 and the first video v2 is greater than the similarity threshold, that is, the second video vv2 corresponds to the first video v 2. The second video vv2 is inserted between the first videos v2 and v3 and at the same time, the first video v20 is deleted to ensure that the new video recommendation still contains 20 videos.
In an exemplary embodiment of the present application, after the first video and the second video are acquired, on-demand states of the first video and the second video may be further acquired. The on-demand status represents the on-demand status of the target user for the first video and the second video, and in practical applications, if the first video and the second video are both training videos, the on-demand status may include, but is not limited to: and waiting for review after the playing is finished, continuing playing after the playing is not finished, waiting for playing after the playing is not played, and the like. The on-demand status of the first video and the second video may be displayed at preset positions of the first video and the second video, for example, at the upper right corners of the first video and the second video, so as to remind the target user to continue playing or review, etc. Meanwhile, the first video and the second video can be classified according to the on-demand state.
In an exemplary embodiment of the present application, in the process of executing step 102, if the vod record of the target user is obtained according to the identification information of the target user, the vod record of the target user is preprocessed, and then the vod record of the target user after the preprocessing operation is used to generate the vod sequence data. And if the video-on-demand record of the target user is not searched according to the identification information of the target user, using preset default video sequence data as the video sequence data of the target user.
In an exemplary embodiment of the present application, the server may store therein default video sequence data, and the default video sequence data may be generated by: and acquiring video-on-demand records of all users, performing the preprocessing operation on the video-on-demand records of all users, generating each video sequence data of all users according to the video-on-demand records of all users after the preprocessing operation, and taking the video sequence data with the highest frequency in each video sequence data of all users as default video sequence data.
In an exemplary embodiment of the present application, the preprocessing operation may include, but is not limited to: deduplication operations, delete missing value operations, and select valid on-demand records, among others.
In an exemplary embodiment of the present application, a training process of a video recommendation model may further be included, where the training process of the video recommendation model may include: counting the on-demand times of all videos in the video-on-demand records of all users, increasing the on-demand times of the videos of which the on-demand times are smaller than a preset time threshold value to obtain a new video-on-demand record, and training according to the new video-on-demand record to obtain a video recommendation model. In order to improve the recommendation accuracy of the video recommendation model, a plurality of latest video-on-demand records of the user can be selected. If a full volume of vod recordings are selected, it is possible that the full volume of vod recordings may not accurately represent the user's on-demand interest. And increasing the on-demand times of the videos with the on-demand times smaller than the time threshold value, namely increasing the weight of the cold videos in the training data. And the situation that part of the competitive cold videos cannot be recommended to the target user is avoided.
Based on the above-described description about the embodiment of the recommendation method for a video, a recommendation method for a training video is described below. The recommendation of the training videos not only needs to recommend the training videos which are watched by other users with similar on-demand modes to the target user, but also needs to recommend new training videos which are not watched by other users to the target user. Not only the training videos which are not watched by the target user are recommended, but also the training videos which are watched are pushed to the target user for reviewing, and the learning effect is strengthened. The relevant videos in the field where the watched training videos are located need to be recommended to the target user, and relevant videos in other fields need to be recommended, so that the learning breadth is improved.
Referring to fig. 2, a flowchart illustrating steps of an embodiment of a method for recommending training videos of the present application is shown. The recommendation method of the training video can be applied to a server, and specifically comprises the following steps:
And performing duplication removal, deletion of missing values, selection of valid play records and the like on the on-demand records read from the database. Also, the most occurring on-demand sequence may be used as the default video sequence data. In practical application, the whole on-demand record of the training videos can be read from the database, test data and off-shelf training videos are excluded according to the data characteristics of the training data, such as sequence length distribution, whether elements of each sequence are consistent and the like, the same user on-demand record of the same training video within 3 minutes is subjected to de-duplication, null value deletion, training video numbering increase and the like, and the effective on-demand sequence of each user is obtained. And selecting the effective on-demand sequence with the highest frequency as default video sequence data from the effective on-demand sequences of each user.
On-demand records of the new training videos are generally few, are not easy to recommend to target users, and need to be recommended separately for the new training videos. Moreover, because the training videos do not have labels and attributes, the similarity between the new training videos and the old training videos can be calculated by using the titles of the training videos. In practical applications, the training video with the earliest on-demand time in the last 1 week in the effective on-demand sequence can be understood as a new training video, and the other videos can be understood as old training videos. The similarity between the new training video and the old training video can be calculated based on a latent semantic analysis method, the latent semantic analysis method can eliminate the influence of synonyms and polysemons, and the method has good expression in the aspect of finding similar documents and titles.
If the video recommendation model is trained by selecting the full on-demand records of the user, the time span of the full on-demand records is large, and the interest of the user is probably changed, so that the current interest of the user can be better captured by selecting the latest n on-demand records. In practical application, the latest 20 on-demand recording training sequence recommendation models of the user can be selected.
And step 204, enhancing the weight of the cold training video.
In addition to the new training videos, there may be some cool competitive training videos worth recommending to the user. However, these cold competitive training videos appear less in the available on-demand sequences, and data amplification is required to increase the probability of recommending the cold competitive training videos. In practical applications, if the frequency of occurrence of a training video in a valid on-demand sequence is less than 20, the frequency of occurrence of the training video may be increased by 2 times to increase the recommendation weight of the training video.
In practical applications, the sequence recommendation model that can be used on the tensoflow deep learning framework is a SR-GNN graph neural network training video recommendation model. Moreover, the steps 201 to 205 may be performed once a day, that is, the video recommendation model is updated and trained, so that the recommendation accuracy of the video recommendation model can be further improved.
And step 206, monitoring a video recommendation request of a target user.
And monitoring a video recommendation page on a target user access terminal to acquire identification information of the target user.
And step 207, returning a training video recommendation result.
And acquiring the on-demand record of the target user according to the identification information of the target user, inputting the on-demand record to the video recommendation model after the preprocessing operation, and outputting the training video recommendation result. If the on-demand record of the target user is empty, the target user is a new user, the default video sequence data can be used as the video sequence data of the target user, the video sequence data of the target user is input into a video recommendation model, and a training video recommendation result is output.
And step 208, recommending a new training video.
And if the old training videos with the similarity difference with the new training videos are in the training video recommendation results, recommending the new training videos to the target user. In practical application, if an old training video with the similarity difference larger than 1 with a new training video is included in the training video recommendation results, the new training video is inserted one bit behind the old training video, and the last training video in the training video recommendation results is deleted. And recommending the new training video to the target user so as to ensure the freshness of the training video recommendation result.
And classifying the training video recommendation result into three categories of review waiting, continuing learning not waiting and learning not waiting according to the characteristics of fragmented learning so as to help a target user to perform complete and systematic training learning. In practical application, the number of training videos in three categories of review completion, continuation completion and non-study completion can be displayed in the training video recommendation result, and corresponding characters of review completion, continuation completion and non-study completion and the like are displayed at the upper right corner of each training video.
Referring to fig. 3, a schematic flow chart of an embodiment of a training video recommendation method of the present application is shown. The recommendation method of the training video can relate to a server, a terminal and a video database. The server sends a data acquisition request to the video database, and the total on-demand records in the video database are subjected to data preprocessing to obtain an effective on-demand sequence. And enhancing the weight of the cold video in the effective on-demand sequence to obtain training data. And training the video recommendation model by using the training data, and storing the trained video recommendation model in the server. When a target user accesses the video recommendation page, the number of the target user is sent to the video database, the on-demand record or the video sequence data of the target user are obtained from the video database, data preprocessing is carried out on the on-demand record or the video sequence data of the target user, then the on-demand record or the video sequence data after data preprocessing is input into the stored video recommendation model, and a video recommendation result is output. Inserting a new training video into the video recommendation result, updating the video recommendation result, performing differential subdivision on the updated video recommendation result, and returning the video recommendation result after differential subdivision to the target user.
According to the embodiment of the application, the gated graph neural network can be used for training the video recommendation model, and the nearest n on-demand recording training video recommendation models are selected, so that the training data volume of the video recommendation model is reduced, and the training time of the video recommendation model is saved. In addition, a new video is added in the video recommendation result, and the probability of recommending the new video is improved.
According to the method and the device, for a new user, namely a user with empty or less on-demand records, a preset default video sequence data input value video recommendation model is utilized, and a video recommendation result of the new user is output.
Under the training video recommending scene, under the condition of recommending the training videos meeting the user interests, the cool competitive training videos can be recommended, so that the learning surface of the user can be widened, and the service level and the knowledge level of the user can be improved.
According to the embodiment of the application, the video recommendation result is differentiated and subdivided, so that a user can learn videos completely, systematically and differentially by utilizing fragmentary time, and the training video learning effect is improved.
It should be noted that, for simplicity of description, the method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the embodiments are not limited by the order of acts described, as some steps may occur in other orders or concurrently depending on the embodiments. Further, those skilled in the art will also appreciate that the embodiments described in the specification are presently preferred and that no particular act is required of the embodiments of the application.
Referring to fig. 4, a block diagram of an embodiment of a video data recommendation apparatus according to the present application is shown, where the video data recommendation apparatus may specifically include the following modules:
a monitoring module 41 configured to monitor a video data recommendation request, where the video data recommendation request includes identification information of a target user;
an obtaining module 42 configured to obtain video sequence data of the target user according to the identification information, the video sequence data being generated based on a plurality of pieces of video-on-demand records arranged in sequence;
a transmission module 43 configured to input the video sequence data to a video recommendation model, and output a video recommendation result, where the video recommendation result includes a plurality of first videos arranged in sequence;
the obtaining module 42 is further configured to obtain at least one corresponding second video according to the similarity of each first video, where the similarity between the second video and the corresponding first video is greater than a preset similarity threshold, and the earliest on-demand time information of the second video is located after the earliest on-demand time information of the corresponding first video;
an update module 44 configured to insert the second video into the video recommendation result, resulting in a new video recommendation result.
In an exemplary embodiment of the present application, the update module 44 is configured to insert the second video after the position of the corresponding first video in the video recommendation, and delete at least one of the first videos in the video recommendation that is reciprocal;
wherein the number of the deleted first video data is the same as the number of the second video data inserted into the video recommendation result.
In an exemplary embodiment of the present application, the obtaining module 42 is further configured to obtain on-demand statuses of the first video and the second video;
the device further comprises: a presentation module configured to present corresponding on-demand states at preset positions of the first video and the second video, and classify the first video and the second video according to the on-demand states.
In an exemplary embodiment of the present application, the obtaining module 42 includes:
the searching module is configured to search for the video-on-demand record of the target user according to the identification information;
a preprocessing module configured to perform a preprocessing operation on the video-on-demand record of the target user;
a generation module configured to generate the video sequence data according to the video-on-demand record of the target user after the preprocessing operation; or,
the searching module is configured to obtain the video-on-demand record of the target user according to the non-searched identification information;
a determination module configured to use preset default video sequence data as the video sequence data of the target user.
In an exemplary embodiment of the present application, the obtaining module 42 is further configured to obtain video-on-demand records of all users;
the preprocessing module is further configured to perform the preprocessing operation on the video-on-demand records of all the users;
the generation module is further configured to generate each video sequence data of all the users according to the video-on-demand records of all the users after the preprocessing operation;
the determining module is further configured to use, as the default video sequence data, video sequence data that appears most frequently among the video sequence data of all the users.
In an exemplary embodiment of the present application, the apparatus further comprises:
the counting module is configured to count the on-demand times of each video in the video on-demand records of all users;
the increasing module is configured to increase the video-on-demand times of videos of which the video-on-demand times are smaller than a preset time threshold value to obtain a new video-on-demand record;
and the training module is configured to obtain the video recommendation model according to the new video-on-demand record training.
In an exemplary embodiment of the present application, the obtaining module 42 is configured to obtain the latest plurality of vod records of all the users.
For the device embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, refer to the partial description of the method embodiment.
The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
Referring now to FIG. 5, shown is a block diagram of a computer system 500 suitable for use in implementing the electronic device of an embodiment of the present application. The electronic device shown in fig. 5 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present application.
As shown in fig. 5, the computer system 500 includes a Central Processing Unit (CPU)501 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)502 or a program loaded from a storage section 508 into a Random Access Memory (RAM) 503. In the RAM503, various programs and data necessary for the operation of the system 500 are also stored. The CPU501, ROM 502, and RAM503 are connected to each other via a bus 504. An input/output (I/O) interface 505 is also connected to bus 504.
The following components are connected to the I/O interface 505: an input portion 506 including a keyboard, a mouse, and the like; an output portion 507 including a display such as a Liquid Crystal Display (LCD) and a speaker; a storage portion 508 including a hard disk and the like; and a communication section 509 including a network interface card such as a LAN card, a modem, or the like. The communication section 509 performs communication processing via a network such as the internet. The driver 510 is also connected to the I/O interface 505 as necessary. A removable medium 511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 510 as necessary, so that a computer program read out therefrom is mounted into the storage section 508 as necessary.
In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 509, and/or installed from the removable medium 511. The computer program performs the above-described functions defined in the method of the present application when executed by the Central Processing Unit (CPU) 501. It should be noted that the computer readable medium described herein can be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In this application, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The units described in the embodiments of the present application may be implemented by software or hardware. The units described may also be provided in a processor, where the names of the units do not in some cases constitute a limitation of the units themselves.
As another aspect, the present application also provides a computer-readable medium, which may be contained in the apparatus described in the above embodiments; or may be present separately and not assembled into the device. The computer readable medium carries one or more programs which, when executed by the apparatus, cause the apparatus to: monitoring a video data recommendation request, wherein the video data recommendation request contains identification information of a target user; acquiring video sequence data of the target user according to the identification information, wherein the video sequence data is generated based on a plurality of video-on-demand records which are sequentially arranged; inputting the video sequence data into a video recommendation model, and outputting a video recommendation result, wherein the video recommendation result comprises a plurality of first videos which are sequentially arranged; acquiring at least one corresponding second video according to the similarity of each first video, wherein the similarity between the second video and the corresponding first video is greater than a preset similarity threshold, and the earliest on-demand time information of the second video is positioned behind the earliest on-demand time information of the corresponding first video; and inserting the second video into the video recommendation result to obtain a new video recommendation result.
The above description is only a preferred embodiment of the application and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the invention herein disclosed is not limited to the particular combination of features described above, but also encompasses other arrangements formed by any combination of the above features or their equivalents without departing from the spirit of the invention. For example, the above features may be replaced with (but not limited to) features having similar functions disclosed in the present application.
Claims (10)
1. A method for recommending video data, comprising:
monitoring a video data recommendation request, wherein the video data recommendation request contains identification information of a target user;
acquiring video sequence data of the target user according to the identification information, wherein the video sequence data is generated based on a plurality of video-on-demand records which are sequentially arranged;
inputting the video sequence data into a video recommendation model, and outputting a video recommendation result, wherein the video recommendation result comprises a plurality of first videos which are sequentially arranged;
acquiring at least one corresponding second video according to the similarity of each first video, wherein the similarity between the second video and the corresponding first video is greater than a preset similarity threshold, and the earliest on-demand time information of the second video is positioned behind the earliest on-demand time information of the corresponding first video;
and inserting the second video into the video recommendation result to obtain a new video recommendation result.
2. The method of claim 1, wherein inserting the second video into the video recommendation results in a new video recommendation result, comprising:
inserting the second video after the position of the corresponding first video in the video recommendation result, and deleting at least one first video which is reciprocal in the video recommendation result;
wherein the number of the deleted first video data is the same as the number of the second video data inserted into the video recommendation result.
3. The method of claim 1, further comprising:
acquiring the on-demand states of the first video and the second video;
displaying corresponding on-demand states at preset positions of the first video and the second video, and classifying the first video and the second video according to the on-demand states.
4. The method of claim 1, wherein the obtaining video sequence data of the target user according to the identification information comprises:
searching to obtain the video-on-demand record of the target user according to the identification information;
carrying out preprocessing operation on the video-on-demand record of the target user;
generating the video sequence data according to the video-on-demand record of the target user after the preprocessing operation; or,
obtaining the video-on-demand record of the target user according to the identification information which is not searched;
and taking preset default video sequence data as the video sequence data of the target user.
5. The method of claim 4, wherein the default video sequence data is generated by:
acquiring video on demand records of all users;
performing the preprocessing operation on the video-on-demand records of all the users;
generating all video sequence data of all users according to the video on demand records of all users after the preprocessing operation;
and taking the video sequence data with the highest frequency in the video sequence data of all the users as the default video sequence data.
6. The method of claim 1, further comprising:
counting the on-demand times of each video in the video on-demand records of all users;
increasing the video-on-demand times of the videos of which the video-on-demand times are smaller than a preset time threshold value to obtain a new video-on-demand record;
and training according to the new video on demand record to obtain the video recommendation model.
7. The method of claim 5, wherein obtaining the video-on-demand records of all users comprises:
and acquiring the latest multiple video-on-demand records of all the users.
8. An apparatus for recommending video data, comprising:
the monitoring module is configured to monitor a video data recommendation request, and the video data recommendation request contains identification information of a target user;
an acquisition module configured to acquire video sequence data of the target user according to the identification information, the video sequence data being generated based on a plurality of pieces of video-on-demand records arranged in sequence;
the transmission module is configured to input the video sequence data into a video recommendation model and output a video recommendation result, wherein the video recommendation result comprises a plurality of first videos which are sequentially arranged;
the obtaining module is further configured to obtain at least one corresponding second video according to the similarity of each first video, the similarity between the second video and the corresponding first video is greater than a preset similarity threshold, and the earliest on-demand time information of the second video is located after the earliest on-demand time information of the corresponding first video;
and the updating module is configured to insert the second video into the video recommendation result to obtain a new video recommendation result.
9. An electronic device, comprising:
one or more processors;
a storage device having one or more programs stored thereon,
when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1-7.
10. A computer-readable medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011206237.7A CN112487240B (en) | 2020-11-02 | 2020-11-02 | Video data recommendation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011206237.7A CN112487240B (en) | 2020-11-02 | 2020-11-02 | Video data recommendation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112487240A true CN112487240A (en) | 2021-03-12 |
| CN112487240B CN112487240B (en) | 2024-03-15 |
Family
ID=74926452
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011206237.7A Active CN112487240B (en) | 2020-11-02 | 2020-11-02 | Video data recommendation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112487240B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103810162A (en) * | 2012-11-05 | 2014-05-21 | 腾讯科技（深圳）有限公司 | Method and system for recommending network information |
| US20170161391A1 (en) * | 2015-12-08 | 2017-06-08 | Le Holdings (Beijing) Co., Ltd. | Method and electronic device for video recommendation |
| CN109284417A (en) * | 2018-08-27 | 2019-01-29 | 广州飞磨科技有限公司 | Video pushing method, device, computer equipment and storage medium |
| CN109429103A (en) * | 2017-08-25 | 2019-03-05 | Tcl集团股份有限公司 | The method, apparatus and computer readable storage medium of recommendation information, terminal device |
| CN110012356A (en) * | 2019-04-16 | 2019-07-12 | 腾讯科技（深圳）有限公司 | Video recommendation method, device and equipment and computer storage medium |
| CN110020122A (en) * | 2017-10-16 | 2019-07-16 | Tcl集团股份有限公司 | A kind of video recommendation method, system and computer readable storage medium |
| CN111491187A (en) * | 2020-04-15 | 2020-08-04 | 腾讯科技（深圳）有限公司 | Video recommendation method, device, equipment and storage medium |
-
2020
- 2020-11-02 CN CN202011206237.7A patent/CN112487240B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103810162A (en) * | 2012-11-05 | 2014-05-21 | 腾讯科技（深圳）有限公司 | Method and system for recommending network information |
| US20170161391A1 (en) * | 2015-12-08 | 2017-06-08 | Le Holdings (Beijing) Co., Ltd. | Method and electronic device for video recommendation |
| CN109429103A (en) * | 2017-08-25 | 2019-03-05 | Tcl集团股份有限公司 | The method, apparatus and computer readable storage medium of recommendation information, terminal device |
| CN110020122A (en) * | 2017-10-16 | 2019-07-16 | Tcl集团股份有限公司 | A kind of video recommendation method, system and computer readable storage medium |
| CN109284417A (en) * | 2018-08-27 | 2019-01-29 | 广州飞磨科技有限公司 | Video pushing method, device, computer equipment and storage medium |
| CN110012356A (en) * | 2019-04-16 | 2019-07-12 | 腾讯科技（深圳）有限公司 | Video recommendation method, device and equipment and computer storage medium |
| CN111491187A (en) * | 2020-04-15 | 2020-08-04 | 腾讯科技（深圳）有限公司 | Video recommendation method, device, equipment and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 陈亮;汪景福;王娜;李霞;: "基于DNN算法的移动视频推荐策略", 计算机学报, no. 08, 29 February 2016 (2016-02-29) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112487240B (en) | 2024-03-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110378732B (en) | Information display method, information association method, device, equipment and storage medium | |
| US9875441B2 (en) | Question recommending method, apparatus and system | |
| US10187344B2 (en) | Social media influence of geographic locations | |
| CN109635155B (en) | Method and device for pushing video to user, electronic equipment and storage medium | |
| CN103229169B (en) | Content providing and system | |
| CN110941738B (en) | Recommendation method and device, electronic equipment and computer-readable storage medium | |
| CN111444304B (en) | Search ordering method and device | |
| EP4083857A1 (en) | Information prediction model training method and apparatus, information prediction method and apparatus, storage medium, and device | |
| CN113239173B (en) | Question-answer data processing method and device, storage medium and electronic equipment | |
| CN112163614A (en) | Anchor classification method and device, electronic equipment and storage medium | |
| CN112579913A (en) | Video recommendation method, device, equipment and computer-readable storage medium | |
| CN114154013A (en) | Video recommendation method, device, equipment and storage medium | |
| CN117407598A (en) | Live broadcast recommendation method and system | |
| CN115618024A (en) | Multimedia recommendation method and device and electronic equipment | |
| CN114845149B (en) | Video clip method, video recommendation method, device, equipment and medium | |
| CN111127057B (en) | Multi-dimensional user portrait recovery method | |
| CN106168975B (en) | The acquisition methods and device of target user's concentration | |
| CN112487240B (en) | Video data recommendation method and device | |
| CN115017200B (en) | Method and device for sorting search results, electronic equipment and storage medium | |
| CN113392266B (en) | Training and sorting method and device of sorting model, electronic equipment and storage medium | |
| CN112035740B (en) | Project use time length prediction method, device, equipment and storage medium | |
| CN114550157A (en) | Bullet screen gathering identification method and device | |
| CN114139006A (en) | Image pushing method, system, equipment and storage medium based on travel product | |
| CN114357242A (en) | Training evaluation method and device based on recall model, equipment and storage medium | |
| CN113094584A (en) | Method and device for determining recommended learning resources |
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 |