Video recommendation method and device

Abstract

The embodiment of the present disclosure provides a video recommendation method and device. The video recommendation method includes: first, in response to determining the current main user, based on the initial portrait information corresponding to the current main user, obtaining a first recommended video, and then sending the first recommended video to the client so that the client plays the first recommended video, in response to receiving a temporary shielding request for the first recommended video, shielding the first recommended video and the associated video of the first recommended video, and based on the temporary shielding request, obtaining the current portrait information corresponding to the current main user, and based on the current portrait information, obtaining and sending a second recommended video so that the client plays the second recommended video, and finally, in response to detecting the recovery instruction of the initial portrait information, obtaining and sending a third recommended video based on the current portrait information and the initial portrait information, and temporarily shielding the currently playing first recommended video and videos of the same type as the first recommended video based on the temporary shielding request.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the fields of computer technology and Internet technology, and in particular, to a video recommendation method and device. Background Art

[0002] There are many classification recommendation scenarios in the short video business, that is, to increase or decrease the recommendation weight of certain types of videos. Common classification recommendation scenarios: not interested, blocked, when a user clicks not interested / blocked on a video, the recommendation weight of the related videos with the corresponding tag of the video is reduced; attention, when a user expresses interest in a video, the recommendation weight of the related videos with the corresponding tag of the video is increased; youth mode, short video apps are forced to provide youth mode, and only display content suitable for teenagers.

[0003] The existing technology currently is to permanently / long-term adjust the classification weights, that is, after the user clicks "not interested", the videos of this type will be reduced or completely stopped, and will not appear for a long time. In actual use, the following scenarios often occur: short-term changes in user interests, differences in occasions, and device lending. In these scenarios, the user's interests will change in the short term, but the long-term interests will not change. Summary of the invention

[0004] Embodiments of the present disclosure provide a video recommendation method, device, electronic device, and computer-readable medium.

[0005] In a first aspect, an embodiment of the present disclosure provides a video recommendation method, the method comprising: in response to determining a current main user, obtaining a first recommended video based on initial portrait information corresponding to the current main user; sending the first recommended video to a client so that the client plays the first recommended video; in response to receiving a temporary shielding request for the first recommended video, shielding the first recommended video and the associated video of the first recommended video, and obtaining current portrait information corresponding to the current main user based on the temporary shielding request; obtaining and sending a second recommended video based on the current portrait information so that the client plays the second recommended video; in response to detecting an instruction to restore the initial portrait information, obtaining and sending a third recommended video based on the current portrait information and the initial portrait information.

[0006] In some embodiments, in response to determining the current main user, a first recommended video is obtained based on initial portrait information corresponding to the current main user, including: determining the current main user based on the client's login usage information; obtaining the initial portrait information of the current main user based on the current main user's historical behavior data; and obtaining the first recommended video based on the initial portrait information of the current main user.

[0007] In some embodiments, the initial portrait information includes a portrait tag and an interest model; and, based on the historical behavior data of the current main user, the initial portrait information of the current main user is obtained, including: constructing a portrait tag corresponding to the current main user based on the historical behavior data of the current main user; performing feature extraction on the historical behavior data to obtain feature information corresponding to the current main user; based on the portrait tag and the feature information, obtaining an interest model corresponding to the current main user; and determining the portrait tag and interest model corresponding to the current main user as the initial portrait information of the current main user.

[0008] In some embodiments, based on the initial portrait information of the current main user, a first recommended video is obtained, including: based on the portrait label corresponding to the current main user, label filtering the candidate video set to obtain a first candidate video; based on the interest model corresponding to the current main user, interest filtering the first candidate video to obtain a second candidate video; using the interest model to sort the second candidate videos to obtain the first recommended video.

[0009] In some embodiments, the temporary blocking request is generated based on the following steps: in response to receiving a temporary blocking operation of the current primary user on the first recommended video, obtaining a video tag of the first recommended video; and generating a temporary blocking request based on the temporary blocking operation and the video tag.

[0010] In some embodiments, shielding the first recommended video and the associated videos of the first recommended video includes: based on the video tag, adding a video filtering condition corresponding to the video tag; based on the video filtering condition, shielding the first recommended video and the associated videos of the first recommended video.

[0011] In some embodiments, the first recommended video and the associated videos of the first recommended video are shielded, including: copying the initial portrait information to obtain the copied portrait information; based on the video tag, updating the copied portrait information to obtain the filtered portrait information; based on the filtered portrait information, shielding the first recommended video and the associated videos of the first recommended video.

[0012] In some embodiments, in response to detecting an instruction to restore the initial portrait information, a third recommended video is obtained and sent based on the current portrait information and the initial portrait information, including: determining whether the current portrait information is the same as the initial portrait information; in response to determining that the current portrait information is not the same as the initial portrait information, obtaining video playback information within preset conditions; based on the video playback information, obtaining user portrait information corresponding to the current main user, and determining whether the user portrait information is the same as the initial portrait information; in response to determining that the user portrait information is the same as the initial portrait information, obtaining and sending the third recommended video based on the user portrait information and the initial portrait information.

[0013] In some embodiments, determining whether the current portrait information is the same as the initial portrait information includes: based on a temporary shielding request, updating the current portrait information and the initial portrait information to obtain new current portrait information and new initial portrait information; calculating the portrait similarity between the new current portrait information and the new initial portrait information; and based on the portrait similarity, determining whether the current portrait information is the same as the initial portrait information.

[0014] In some embodiments, calculating the portrait similarity between new current portrait information and new initial portrait information includes: performing feature extraction on the new current portrait information and the new initial portrait information respectively to obtain first feature information and second feature information; preprocessing the first feature information and the second feature information to obtain processed first feature information and processed second feature information; calculating the feature similarity between the processed first feature information and the processed second feature information; and calculating the portrait similarity between the new current portrait information and the new initial portrait information based on the feature similarity and the weight value corresponding to the feature similarity.

[0015] In some embodiments, in response to determining that the user portrait information is the same as the initial portrait information, based on the user portrait information and the initial portrait information, obtaining and sending a third recommended video, including: in response to determining that the user portrait information is the same as the initial portrait information, respectively extracting features from the user portrait information and the initial portrait information to obtain third feature information and fourth feature information; preprocessing the third feature information and the fourth feature information to obtain processed third feature information and processed fourth feature information; based on weight values ​​corresponding to the processed third feature information and the processed fourth feature information, performing feature fusion on the processed third feature information and the processed fourth feature information to obtain fused feature information; based on the fused feature information, obtaining fused portrait information corresponding to the current main user; based on the fused portrait information, obtaining and sending the third recommended video.

[0016] In some embodiments, in response to detecting an instruction to restore the initial portrait information, obtaining and sending a third recommended video based on the current portrait information and the initial portrait information, also includes: in response to determining that the current portrait information is the same as the initial portrait information, obtaining and sending the first video corresponding to the initial portrait information; based on the operation information of the current main user on the first video, determining the portrait information used to push the video from the current portrait information and the initial portrait information; and obtaining and sending the third recommended video based on the portrait information.

[0017] In some embodiments, based on the operation information of the current main user on the first video, the portrait information used for pushing the video is determined from the current portrait information and the initial portrait information, including: in response to determining that the operation information indicates that the first video is playing normally, shortening the push time interval of the first video; based on the shortened push time interval, sending the second video corresponding to the initial portrait information; obtaining the statistical number of operation information indicating normal playback; in response to determining that the statistical number exceeds a preset number, obtaining the fused portrait information corresponding to the current main user based on the user portrait information and the initial portrait information, and determining the fused portrait information as the portrait information used for pushing the video.

[0018] In some embodiments, based on the operation information of the current main user on the first video, the portrait information used for pushing the video is determined from the current portrait information and the initial portrait information, and also includes: in response to determining that the operation information indicates that the first video is temporarily blocked, executing a video detection step: based on the last push time and the push time interval of the first video, determining the current push time of the first video; based on the current push time, sending the first video corresponding to the initial portrait information; obtaining the operation information of the current main user on the first video; until it is determined that the operation information indicates that the first video is playing normally, determining the portrait information used for pushing the video from the current portrait information and the initial portrait information.

[0019] In a second aspect, an embodiment of the present disclosure provides a video recommendation device, which includes: a first acquisition module, configured to acquire a first recommended video based on initial portrait information corresponding to the current main user in response to determining a current main user; a sending module, configured to send the first recommended video to a client, so that the client plays the first recommended video; a second acquisition module, configured to shield the first recommended video and the associated video of the first recommended video in response to receiving a temporary shielding request for the first recommended video, and acquire current portrait information corresponding to the current main user based on the temporary shielding request; a third acquisition module, configured to acquire and send a second recommended video based on the current portrait information, so that the client plays the second recommended video; a fourth acquisition module, configured to acquire and send a third recommended video based on the current portrait information and the initial portrait information in response to detecting an instruction to restore the initial portrait information.

[0020] In some embodiments, the first acquisition module is further configured to: determine the current main user based on the client's login usage information; obtain the initial portrait information of the current main user based on the historical behavior data of the current main user; and obtain the first recommended video based on the initial portrait information of the current main user.

[0021] In some embodiments, the initial portrait information includes a portrait tag and an interest model; and the first acquisition module is further configured to: construct a portrait tag corresponding to the current main user based on the historical behavior data of the current main user; perform feature extraction on the historical behavior data to obtain feature information corresponding to the current main user; based on the portrait tag and feature information, obtain the interest model corresponding to the current main user; determine the portrait tag and interest model corresponding to the current main user as the initial portrait information of the current main user.

[0022] In some embodiments, the first acquisition module is further configured to: perform label screening on the candidate video set based on the portrait label corresponding to the current main user to obtain a first candidate video; perform interest screening on the first candidate video based on the interest model corresponding to the current main user to obtain a second candidate video; and use the interest model to sort the second candidate video to obtain a first recommended video.

[0023] In some embodiments, the temporary blocking request is generated based on the following steps: in response to receiving a temporary blocking operation of the current primary user on the first recommended video, obtaining a video tag of the first recommended video; and generating a temporary blocking request based on the temporary blocking operation and the video tag.

[0024] In some embodiments, the second acquisition module is further configured to: based on the video tag, add a video filtering condition corresponding to the video tag; based on the video filtering condition, shield the first recommended video and the associated videos of the first recommended video.

[0025] In some embodiments, the second acquisition module is further configured to: copy the initial portrait information to obtain copied portrait information; update the copied portrait information based on the video tag to obtain filtered portrait information; and shield the first recommended video and the associated videos of the first recommended video based on the filtered portrait information.

[0026] In some embodiments, the fourth acquisition module is further configured to: determine whether the current portrait information is the same as the initial portrait information; in response to determining that the current portrait information is not the same as the initial portrait information, obtain video playback information within preset conditions; based on the video playback information, obtain user portrait information corresponding to the current main user, and determine whether the user portrait information is the same as the initial portrait information; in response to determining that the user portrait information is the same as the initial portrait information, obtain and send a third recommended video based on the user portrait information and the initial portrait information.

[0027] In some embodiments, the fourth acquisition module is further configured to: based on a temporary shielding request, update the current portrait information and the initial portrait information to obtain new current portrait information and new initial portrait information; calculate the portrait similarity between the new current portrait information and the new initial portrait information; based on the portrait similarity, determine whether the current portrait information is the same as the initial portrait information.

[0028] In some embodiments, the fourth acquisition module is further configured to: perform feature extraction on the new current portrait information and the new initial portrait information, respectively, to obtain first feature information and second feature information; preprocess the first feature information and the second feature information, to obtain processed first feature information and processed second feature information; calculate the feature similarity between the processed first feature information and the processed second feature information; and calculate the portrait similarity between the new current portrait information and the new initial portrait information based on the feature similarity and the weight value corresponding to the feature similarity.

[0029] In some embodiments, the fourth acquisition module is further configured to: in response to determining that the user portrait information is the same as the initial portrait information, respectively extract features from the user portrait information and the initial portrait information to obtain third feature information and fourth feature information; pre-process the third feature information and the fourth feature information to obtain processed third feature information and processed fourth feature information; based on the weight values ​​corresponding to the processed third feature information and the processed fourth feature information, perform feature fusion on the processed third feature information and the processed fourth feature information to obtain fused feature information; based on the fused feature information, obtain fused portrait information corresponding to the current main user; based on the fused portrait information, obtain and send the third recommended video.

[0030] In some embodiments, the fourth acquisition module is further configured to: in response to determining that the current portrait information is the same as the initial portrait information, acquire and send the first video corresponding to the initial portrait information; based on the operation information of the current main user on the first video, determine the portrait information used to push the video from the current portrait information and the initial portrait information; based on the portrait information, acquire and send the third recommended video.

[0031] In some embodiments, the fourth acquisition module is further configured to: in response to determining that the operation information indicates that the first video is playing normally, shorten the push time interval of the first video; based on the shortened push time interval, send the second video corresponding to the initial portrait information; obtain the statistical number of operation information indicating normal playback; in response to determining that the statistical number exceeds a preset number, based on the user portrait information and the initial portrait information, obtain the fused portrait information corresponding to the current main user, and determine the fused portrait information as the portrait information used to push the video.

[0032] In some embodiments, the fourth acquisition module is further configured to: in response to determining that the operation information indicates that the first video is temporarily blocked, execute the video detection step: determine the current push time of the first video based on the last push time and the push time interval of the first video; based on the current push time, send the first video corresponding to the initial portrait information; obtain the operation information of the current main user on the first video; until it is determined that the operation information indicates that the first video is playing normally, determine the portrait information used to push the video from the current portrait information and the initial portrait information.

[0033] In a third aspect, an embodiment of the present disclosure provides an electronic device, comprising: one or more processors; a storage device on which one or more programs are stored; when the one or more programs are executed by one or more processors, the one or more processors implement the video recommendation method described in any embodiment of the first aspect.

[0034] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable medium having a computer program stored thereon, and when the computer program is executed by a processor, the video recommendation method as described in any embodiment of the first aspect is implemented.

[0035] The video recommendation method provided by the embodiment of the present disclosure comprises the following steps: the execution subject first responds to determining the current main user, obtains a first recommended video based on the initial portrait information corresponding to the current main user, and then sends the first recommended video to the client so that the client plays the first recommended video; thereafter, in response to receiving a temporary shielding request for the first recommended video, shields the first recommended video and the associated video of the first recommended video, obtains the current portrait information corresponding to the current main user based on the temporary shielding request, obtains and sends the second recommended video based on the current portrait information, so that the client plays the second recommended video; and finally, in response to detecting a recovery instruction for the initial portrait information, obtains and sends the third recommended video based on the current portrait information and the initial portrait information, and then plays the video. During the video recommendation process, a temporary video shielding function can be provided. Based on the temporary shielding request, the first recommended video currently being played and videos of the same type as the first recommended video can be temporarily shielded. The appearance of a certain type of short video can be restricted or the recommendation weight of this type of short video can be reduced within a short period of time. After that, when the instruction to restore the initial portrait information is detected, the third recommended video can be obtained according to the user portrait information and the initial portrait information. That is, after the temporary restriction is lifted, the original short video preference or recommendation weight is restored, which realizes the introduction of the temporary shielding function, providing users with the ability to temporarily shield certain short video categories. After the conditions for restoring the initial portrait information are met, the short video preference recommendation corresponding to the initial portrait information is automatically restored, which improves the flexibility of video recommendation and the targeted video recommendation for different scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Other features, objects and advantages of the present disclosure will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0037] Figure 1 is an exemplary system architecture diagram in which an embodiment of the present disclosure may be applied;

[0038] Figure 2 is a flowchart of an embodiment of a video recommendation method according to the present disclosure;

[0039] Figure 3 is a flowchart of an embodiment of obtaining a first recommended video according to the present disclosure;

[0040] Figure 4 is a flowchart of another embodiment of the video recommendation method according to the present disclosure;

[0041] Figure 5 is a flow chart of an embodiment of determining whether current portrait information is the same as initial portrait information according to the present disclosure;

[0042] Figure 6 is a flow chart of an embodiment of obtaining and sending a third recommended video according to the present disclosure;

[0043] Figure 7 It is a flowchart of an embodiment of determining the portrait information for pushing the video from the current portrait information and the initial portrait information based on the operation information of the current main user on the first video according to the present disclosure;

[0044] Figure 8 is a structural schematic diagram of an embodiment of a video recommendation device according to the present disclosure;

[0045] Fig. 9 It is a schematic diagram of the structure of an electronic device suitable for implementing the embodiments of the present disclosure. DETAILED DESCRIPTION

[0046] The present disclosure is further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant disclosure, rather than to limit the disclosure. It is also necessary to explain that, for ease of description, only the parts related to the relevant disclosure are shown in the accompanying drawings.

[0047] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present disclosure may be combined with each other. The present disclosure will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0048] Figure 1 An exemplary system architecture 100 is shown to which the video recommendation method and apparatus according to the embodiments of the present disclosure can be applied.

[0049] like Figure 1 As shown, the system architecture 100 may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 is used to provide a medium for communication links between the terminal devices 101, 102, 103 and the server 105. The network 104 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0050] Users can use terminal devices 101, 102, 103 to interact with server 105 through network 104 to receive or send messages, etc. Terminal devices 101, 102, 103 can be user terminal devices, on which various client applications can be installed, such as image applications, video applications, search applications, financial applications, etc.

[0051] Terminal devices 101, 102, 103 can be various electronic devices with display screens and supporting receiving server messages, including but not limited to smart phones, tablet computers, e-book readers, electronic players, laptop computers and desktop computers, etc., and can be devices with image acquisition devices for performing image acquisition.

[0052] Terminal devices 101, 102, 103 can be hardware or software. When terminal devices 101, 102, 103 are hardware, they can be various electronic devices. When terminal devices 101, 102, 103 are software, they can be installed in the electronic devices listed above. They can be implemented as multiple software or software modules (for example, multiple software modules for providing distributed services), or they can be implemented as a single software or software module. No specific limitation is made here.

[0053] The server 105 may be a server providing various services, such as a backend server that receives a request sent by a terminal device that establishes a communication connection with the server 105. The backend server may receive and analyze the request sent by the terminal device and generate a processing result.

[0054] Server 105 can determine the current main user, obtain the first recommended video based on the initial portrait information corresponding to the current main user, and then send the first recommended video to the client so that the client plays the first recommended video. If a temporary blocking request for the first recommended video is received, the first recommended video and the associated video of the first recommended video are blocked, and based on the temporary blocking request, the current portrait information corresponding to the current main user is obtained, and based on the current portrait information, the second recommended video is obtained and sent, so that the client plays the second recommended video. Finally, if an instruction to restore the initial portrait information is detected, the third recommended video is obtained and sent based on the current portrait information and the initial portrait information.

[0055] It should be noted that the server can be hardware or software. When the server is hardware, it can be various electronic devices that provide various services to the terminal device. When the server is software, it can be implemented as multiple software or software modules that provide various services to the terminal device, or it can be implemented as a single software or software module that provides various services to the terminal device. No specific limitation is made here.

[0056] It should be noted that the video recommendation method provided in the embodiments of the present disclosure may be executed by the server 105 , and accordingly, the video recommendation device may be disposed in the server 105 .

[0057] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0058] refer to Figure 2 , a flowchart 200 of an embodiment of a video recommendation method according to the present disclosure is shown. The video recommendation method comprises the following steps:

[0059] Step 210, in response to determining the current main user, obtaining a first recommended video based on initial portrait information corresponding to the current main user.

[0060] In this step, the execution subject (e.g. Figure 1 The server 105 in the example can determine the current main user corresponding to the client according to the usage information of the client. The current main user can represent the user who controls the client, that is, the user who temporarily restricts the short video preference by operating the temporary shielding function. The above-mentioned execution subject can read the initial portrait information corresponding to the current main user through the network. The initial portrait information can represent the complete portrait information of the current main user.

[0061] The above-mentioned execution entity can use the initial portrait information of the current main user to screen all candidate videos, screen out multiple candidate videos corresponding to the initial portrait information, and determine them as the first recommended video. The first recommended video may include multiple videos to be recommended corresponding to the initial portrait information.

[0062] Step 220: Send the first recommended video to the client, so that the client plays the first recommended video.

[0063] In this step, after the execution subject obtains the first recommended video corresponding to the initial portrait information, the first recommended video can be sent to the client. After the client receives the first recommended video, the first recommended video can be played to the current main user through the screen.

[0064] Step 230, in response to receiving a temporary shielding request for the first recommended video, shielding the first recommended video and the associated videos of the first recommended video, and based on the temporary shielding request, obtaining current portrait information corresponding to the current main user.

[0065] In this step, the current main user can perform a temporary shielding operation on the first recommended video during the playback of the first recommended video. The temporary shielding operation can be that the current main user presents a temporary shielding control by long pressing the current playback interface and selects the temporary shielding control, or the temporary shielding operation can be that the current main user selects the temporary shielding control that has been presented in the current playback interface, or the temporary shielding operation can be that the current main user clicks or enters the setting interface and selects the temporary shielding control in the setting interface. The client generates a temporary shielding request for the first recommended video according to the temporary shielding operation, and sends the temporary shielding request to the above-mentioned execution subject. After receiving the temporary shielding request for the first recommended video, the above-mentioned execution subject can obtain the associated video associated with the first recommended video, which may include a video of the same video type as the first recommended video, and shield the first recommended video and the associated video of the first recommended video, temporarily shielding the first recommended video and the associated video of the first recommended video, and then update the initial portrait information of the current main user according to the temporary shielding request to obtain the current portrait information corresponding to the current main user.

[0066] As an optional implementation, the temporary blocking request may be generated based on the following steps: in response to receiving a temporary blocking operation of the current main user on the first recommended video, obtaining a video tag of the first recommended video; and generating a temporary blocking request based on the temporary blocking operation and the video tag.

[0067] Specifically, after receiving the temporary blocking operation of the current main user on the first recommended video, the client can obtain the video tag of the first recommended video corresponding to the temporary blocking operation. Then, the client generates a temporary blocking request corresponding to the first recommended video according to the video tag and the temporary blocking operation, and sends the temporary blocking request to the above-mentioned execution subject.

[0068] In this implementation, by generating a temporary blocking request based on the temporary blocking operation and the video tag, the temporary blocking request corresponding to the blocked video can be accurately determined.

[0069] As an optional implementation method, the shielding processing of the first recommended video and the associated videos of the first recommended video in the above step 230 may include the following steps: based on the video tag, adding a video filtering condition corresponding to the video tag; based on the video filtering condition, shielding the first recommended video and the associated videos of the first recommended video.

[0070] Specifically, after the execution subject obtains the temporary blocking request, it can analyze the video tag in the temporary blocking request, and add a video filtering condition corresponding to the video tag according to the video tag. Then, the execution subject filters the first recommended video and the associated videos of the first recommended video according to the video filtering condition, and temporarily blocks the first recommended video corresponding to the video tag.

[0071] As an example, the execution entity may add the video tag to a filter layer, and when sorting the video recall, add the filter layer to filter out the related videos in the recall results, and return the filtered videos.

[0072] In this implementation, by adding filtering conditions, the first recommended video and the associated videos of the first recommended video are screened and filtered, so that the first recommended video and the associated videos corresponding to the video tag can be temporarily shielded, thereby improving the accuracy of video shielding.

[0073] Alternatively, as an optional implementation method, the shielding processing of the first recommended video and the associated videos of the first recommended video in the above step 230 may include the following steps: copying the initial portrait information to obtain the copied portrait information; based on the video tag, updating the copied portrait information to obtain the filtered portrait information; based on the filtered portrait information, shielding the first recommended video and the associated videos of the first recommended video.

[0074] Specifically, the execution subject may copy the initial portrait information to obtain the copied portrait information, and store the initial portrait information. The execution subject may update the copied portrait information according to the video tag, delete the corresponding video tag in the copied portrait information, and obtain the filtered portrait information. Then, the execution subject may use the filtered portrait information to shield the first recommended video and the associated video of the first recommended video, temporarily shield the first recommended video corresponding to the video tag and the associated video of the first recommended video, and obtain a new recommended video.

[0075] In this implementation, by updating and copying the portrait information, the first recommended video and the associated videos of the first recommended video are shielded by using the filtered portrait information, so that the first recommended video and the associated videos corresponding to the video tag can be temporarily shielded, thereby improving the accuracy of video shielding.

[0076] Step 240, based on the current portrait information, obtain and send the second recommended video so that the client plays the second recommended video.

[0077] In this step, after the above-mentioned execution subject obtains the current portrait information corresponding to the current main user, it can use the current portrait information to screen all candidate videos, screen out multiple candidate videos corresponding to the current portrait information, and determine them as the second recommended video. The second recommended video may include multiple videos to be recommended corresponding to the current portrait information.

[0078] Step 250, in response to detecting an instruction to restore the initial portrait information, obtaining and sending a third recommended video based on the current portrait information and the initial portrait information.

[0079] In this step, the execution subject can detect the recovery instruction of the initial portrait information in real time when using the current user portrait information to recommend videos. For example, it can detect whether the video corresponding to the video tag within the preset time is played normally, whether the number of played videos meets the preset number, whether the current main user is watching the temporarily blocked video, etc. After the execution subject determines that the recovery condition of the initial portrait information is met through detection, it can use the current portrait information and the initial portrait information at the same time to screen all candidate videos, obtain the third recommended video, and send the third recommended video to the client so that the client can play the third recommended video through the screen.

[0080] The video recommendation method provided by the embodiment of the present disclosure comprises the following steps: the execution subject first responds to determining the current main user, obtains a first recommended video based on the initial portrait information corresponding to the current main user, and then sends the first recommended video to the client so that the client plays the first recommended video; thereafter, in response to receiving a temporary shielding request for the first recommended video, shields the first recommended video and the associated video of the first recommended video, obtains the current portrait information corresponding to the current main user based on the temporary shielding request, obtains and sends the second recommended video based on the current portrait information, so that the client plays the second recommended video; and finally, in response to detecting a recovery instruction for the initial portrait information, obtains and sends the third recommended video based on the current portrait information and the initial portrait information, and then plays the video. During the video recommendation process, a temporary video shielding function can be provided. Based on the temporary shielding request, the first recommended video currently being played and videos of the same type as the first recommended video can be temporarily shielded. The appearance of a certain type of short video can be restricted or the recommendation weight of this type of short video can be reduced within a short period of time. After that, when the instruction to restore the initial portrait information is detected, the third recommended video can be obtained according to the user portrait information and the initial portrait information. That is, after the temporary restriction is lifted, the original short video preference or recommendation weight is restored, which realizes the introduction of the temporary shielding function, providing users with the ability to temporarily shield certain short video categories. After the conditions for restoring the initial portrait information are met, the short video preference recommendation corresponding to the initial portrait information is automatically restored, which improves the flexibility of video recommendation and the targeted video recommendation for different scenarios.

[0081] See also Figure 3 , Figure 3 A flowchart 300 of an embodiment of obtaining a first recommended video is shown, that is, the above step 210, in response to determining the current main user, obtaining the first recommended video based on the initial portrait information corresponding to the current main user, which may include the following steps:

[0082] Step 310, determining the current primary user based on the client's login usage information.

[0083] In this step, the execution subject may obtain the login and use information of the client, which may represent the login status information and use information of the client. The execution subject may analyze the login and use information of the client, and determine the current primary user according to the login and use information of the client.

[0084] As an example, the execution subject analyzes the main login user of the client according to the login usage information, and determines the main login user as the current main user.

[0085] As an example, the execution subject analyzes the main visitor user when the client is not logged in according to the login usage information, and determines the main visitor user as the current main user.

[0086] As an example, the above-mentioned execution subject analyzes the user who has continuously used the video APP in a recent period of time in certain scenarios of the client according to the login usage information, and determines the user as the current main user.

[0087] As an example, the execution subject analyzes the user who operates the temporary shielding function on the client according to the login usage information, and determines the user as the current main user.

[0088] Step 320, based on the historical behavior data of the current main user, obtain the initial portrait information of the current main user.

[0089] In this step, the execution subject can read the historical behavior data of the current main user through the network, and the historical behavior data may include the viewing history, likes, comments, shares, searches, and attention of the current main user. Then the execution subject performs data analysis on the historical behavior data of the current main user, constructs a user portrait, analyzes the interest tags, preference types, viewing habits, and other information of the current main user, and obtains the initial portrait information of the current main user.

[0090] As an optional implementation, the initial portrait information may include a portrait tag and an interest model; and the step 320 of obtaining the initial portrait information of the current primary user based on the historical behavior data of the current primary user may include the following steps:

[0091] The first step is to build a portrait label corresponding to the current main user based on the historical behavior data of the current main user.

[0092] Specifically, the above-mentioned execution entity can perform data analysis on the historical behavior data of the current main user, determine the interest tags, preference types, viewing habits and other tag information of the current main user, and construct a portrait tag corresponding to the current main user based on the determined tag information.

[0093] The second step is to extract features from the historical behavior data to obtain feature information corresponding to the current main user.

[0094] Specifically, the above-mentioned execution entity can use statistical methods or machine learning methods to extract features from historical behavior data, extract meaningful features, such as number of views, like ratio, focus areas, etc., and obtain feature information corresponding to the current main user.

[0095] The third step is to obtain the interest model corresponding to the current main user based on the portrait label and feature information.

[0096] Specifically, after the above-mentioned execution entity obtains the portrait label and feature information corresponding to the current main user, it can use machine learning algorithms, such as collaborative filtering, deep learning models and other algorithms, to establish an interest model corresponding to the current main user based on the portrait label and feature information corresponding to the current main user, so that the interest model can predict the current main user's interest in short videos of different labels or types.

[0097] The fourth step is to determine the portrait label and interest model corresponding to the current main user as the initial portrait information of the current main user.

[0098] Specifically, after the above-mentioned execution subject obtains the interest model corresponding to the current main user, it can determine the portrait label and interest model corresponding to the current main user as the initial portrait information of the current main user.

[0099] In this implementation, the portrait label is determined by utilizing the historical behavior data of the current main user, and features are extracted from the historical behavior data to construct an interest model using the feature information. This makes the initial portrait information of the current main user include both the portrait label and the interest model, thereby improving the comprehensiveness and accuracy of the initial portrait information.

[0100] It should be noted that when temporary shielding takes effect, the initial portrait information is copied, one of which remains unchanged (for subsequent recovery), and the other is updated after removing the shielded content as a temporary portrait.

[0101] Step 330: Obtain a first recommended video based on the initial portrait information of the current main user.

[0102] In this step, the above-mentioned execution entity can use the initial portrait information of the current main user to screen all candidate videos, screen out multiple candidate videos corresponding to the initial portrait information, and determine them as the first recommended video. The first recommended video may include multiple videos to be recommended corresponding to the initial portrait information.

[0103] As an optional implementation, the initial portrait information may include a portrait tag and an interest model; and the step 330 of obtaining the first recommended video based on the initial portrait information of the current primary user may include the following steps:

[0104] In the first step, based on the portrait label corresponding to the current main user, the candidate video set is labeled and filtered to obtain the first candidate video.

[0105] Specifically, the above-mentioned execution entity can use a recommendation algorithm according to the portrait label corresponding to the current main user, such as content-based recommendation, collaborative filtering, deep learning recommendation model, etc., to perform label screening on the candidate video set and obtain a first candidate video consisting of multiple candidate videos that meet the portrait label.

[0106] In the second step, based on the interest model corresponding to the current main user, the first candidate video is screened for interest to obtain a second candidate video.

[0107] Specifically, the execution subject uses the interest model corresponding to the current main user to screen the interest of the first candidate video, predicts the current main user's prediction for the first candidate video, and determines the second candidate video from the first candidate video.

[0108] The third step is to use the interest model to sort the second candidate videos to obtain the first recommended video.

[0109] Specifically, the above-mentioned execution entity can use the interest model to sort the second candidate videos, that is, use the interest model to sort the predicted interest values ​​of the second candidate videos, and screen out multiple candidate videos whose predicted interest values ​​exceed a threshold, thereby obtaining the first recommended video.

[0110] In this implementation, the candidate video collection is screened through portrait tags and interest models, and a combination of different algorithms is used to recommend short videos that match the user's interests, making video screening more precise and improving the accuracy of the first recommended video.

[0111] In this implementation, the first recommended video is obtained by utilizing the initial portrait information corresponding to the current main user, so that video screening is more precise and the accuracy of the first recommended video is improved.

[0112] refer to Figure 4 , Figure 4A flowchart 400 of another embodiment of a video recommendation method is shown. The video recommendation method may include the following steps:

[0113] Step 410, in response to determining the current main user, obtaining a first recommended video based on initial portrait information corresponding to the current main user.

[0114] In this step, step 410 and Figure 2 Step 210 in the illustrated embodiment is the same and will not be described in detail here.

[0115] Step 420: Send the first recommended video to the client, so that the client plays the first recommended video.

[0116] In this step, step 420 and Figure 2 Step 220 in the illustrated embodiment is the same and will not be described in detail here.

[0117] Step 430, in response to receiving a temporary shielding request for the first recommended video, shielding the first recommended video and the associated videos of the first recommended video, and based on the temporary shielding request, obtaining current portrait information corresponding to the current main user.

[0118] In this step, step 430 and Figure 2 Step 230 in the illustrated embodiment is the same and will not be described in detail here.

[0119] Step 440, based on the current portrait information, obtain and send the second recommended video so that the client plays the second recommended video.

[0120] In this step, step 440 and Figure 2 Step 240 in the illustrated embodiment is the same and will not be described in detail here.

[0121] Step 450, determining whether the current portrait information is the same as the initial portrait information.

[0122] In this step, the above-mentioned execution entity can compare the current portrait information with the initial portrait information, and can compare each feature information in the current portrait information with each feature information in the initial portrait information to determine whether the current portrait information is the same as the initial portrait information.

[0123] Step 460, in response to determining that the current portrait information is different from the initial portrait information, obtaining video playback information within preset conditions.

[0124] In this step, the execution subject determines that the current portrait information is different from the initial portrait information, and then obtains the video playback information within the preset conditions, which may include a preset time period or a preset number of played videos. That is, the execution subject can obtain the video playback information within the preset time period, or the execution subject can obtain the video playback information within the preset number of played videos. The video playback information may be the video that has been played by the client and the video playback status information, and the video playback status information may include normal video playback, temporary video blocking, and paused video playback.

[0125] Step 470, based on the video playback information, obtain the user portrait information corresponding to the current main user, and determine whether the user portrait information is the same as the initial portrait information.

[0126] In this step, the above-mentioned execution entity can analyze the video playback information and obtain the user portrait information corresponding to the current main user. The user portrait information can be a portrait label and interest model corresponding to the video playback information.

[0127] Then the above-mentioned execution entity can compare the user portrait information with the initial portrait information, compare each feature information in the user portrait information with each feature information in the initial portrait information, and determine whether the user portrait information is the same as the initial portrait information.

[0128] Step 480, in response to determining that the user portrait information is the same as the initial portrait information, obtaining and sending a third recommended video based on the user portrait information and the initial portrait information.

[0129] In this step, if the above-mentioned execution entity determines that the user portrait information is the same as the initial portrait information, it can be determined that the user portrait information of the current main user within the preset conditions is close to the initial portrait information of the current main user, the temporary shielding can be lifted, and a prompt message that the temporary shielding has been lifted is sent to the client, so that the client can present the prompt message to the current main user.

[0130] Then, the execution subject can use the current portrait information and the initial portrait information at the same time to screen all candidate videos, obtain the third recommended video, and send the third recommended video to the client so that the client can play the third recommended video through the screen.

[0131] In this embodiment, by determining that the current portrait information is different from the initial portrait information, obtaining the video playback information within the preset conditions, and judging whether the user portrait information corresponding to the video playback information is the same as the initial portrait information, when it is determined that the user portrait information is the same as the initial portrait information, temporary shielding can be lifted, and at the same time, the current portrait information and the initial portrait information are used to obtain the third recommended video, thereby realizing the restoration of the portrait information after the portrait information has changed, and improving the flexibility of video shielding and video recovery.

[0132] refer to Figure 5 , Figure 5 A flowchart 500 is shown of an embodiment of determining whether the current portrait information is the same as the initial portrait information, that is, the above step 450, determining whether the current portrait information is the same as the initial portrait information may include the following steps:

[0133] Step 510, based on the temporary shielding request, the current portrait information and the initial portrait information are updated to obtain new current portrait information and new initial portrait information.

[0134] In this step, the above-mentioned execution entity can update the current portrait information and the initial portrait information according to the temporary shielding request, remove the video tags corresponding to the temporary shielding operation request in the current portrait information and the initial portrait information, and obtain new current portrait information and new initial portrait information, which can reduce interference items in the portrait information.

[0135] Step 520, calculating the portrait similarity between the new current portrait information and the new initial portrait information.

[0136] In this step, the above-mentioned execution entity can compare the new current portrait information with the new initial portrait information, analyze each feature information in the new current portrait information and the new initial portrait information, and use cosine similarity to calculate the portrait similarity between the new current portrait information and the new initial portrait information.

[0137] As an optional implementation, the above step 520, calculating the portrait similarity between the new current portrait information and the new initial portrait information, may include the following steps:

[0138] In the first step, feature extraction is performed on the new current portrait information and the new initial portrait information respectively to obtain the first feature information and the second feature information.

[0139] Specifically, the above-mentioned execution entity can perform feature extraction on the new current portrait information and the new initial portrait information respectively, extract meaningful features from the new current portrait information and the new initial portrait information, which may include the user's behavior, interests, preferences, etc., and extract the same feature information from the new current portrait information and the new initial portrait information to obtain the first feature information and the second feature information.

[0140] In the second step, the first feature information and the second feature information are preprocessed to obtain processed first feature information and processed second feature information.

[0141] Specifically, the above-mentioned execution entity can preprocess the first feature information and the second feature information, and the preprocessing may include missing value processing, standardization and other processing operations, so that the first feature information and the second feature information are in a comparable state, and the processed first feature information and the processed second feature information are obtained.

[0142] The third step is to calculate the feature similarity between the processed first feature information and the processed second feature information.

[0143] Specifically, the execution subject may use cosine similarity to calculate the feature similarity between the processed first feature information and the processed second feature information. The calculation formula of the cosine similarity is:

[0144] similarity=(A·B) / (||A||*||B||)

[0145] Among them, A and B are two vectors, respectively representing the values ​​of the processed first feature information and the processed second feature information in the new current portrait information and the new initial portrait information.

[0146] The fourth step is to calculate the portrait similarity between the new current portrait information and the new initial portrait information based on the feature similarity and the weight value corresponding to the feature similarity.

[0147] Specifically, the above-mentioned execution entity can obtain the weight value corresponding to the feature similarity, and the weight value can be determined according to business needs and data analysis. The above-mentioned execution entity can calculate the portrait similarity between the new current portrait information and the new initial portrait information based on the feature similarity and the weight value corresponding to the feature similarity. For example, for each feature similarity, a weighted sum can be performed according to the weight value of the feature similarity to obtain the portrait similarity between the new current portrait information and the new initial portrait information. Then the above-mentioned execution entity can also normalize the calculated portrait similarity to ensure that it is within a unified range for easy comparison and analysis.

[0148] In this implementation, by using cosine similarity as a similarity calculation method, cosine similarity can handle the weight problem between different features, so that the calculation of portrait similarity is more accurate, thereby improving the accuracy of portrait similarity.

[0149] Step 530, based on the portrait similarity, determine whether the current portrait information is the same as the initial portrait information.

[0150] In this step, after the execution subject obtains the portrait similarity, it can compare the portrait similarity with a preset threshold, which can be a similarity threshold pre-set according to business needs. The execution subject can determine whether the portrait similarity is greater than the preset threshold, thereby determining whether the current portrait information is the same as the initial portrait information.

[0151] If it is determined that the portrait similarity is greater than or equal to the preset threshold, it is determined that the current portrait information is the same as the initial portrait information; if it is determined that the portrait similarity is less than the preset threshold, it is determined that the current portrait information is different from the initial portrait information

[0152] In this implementation, the current portrait information and the initial portrait information are updated to obtain new current portrait information and new initial portrait information, so as to calculate the portrait similarity between the new current portrait information and the new initial portrait information, thereby determining whether the current portrait information is the same as the initial portrait information. The video label corresponding to the temporary shielding operation can be eliminated, and interference items can be eliminated, making the judgment more accurate and improving the accuracy of the judgment.

[0153] refer to Figure 6 , Figure 6 A flowchart 600 of an embodiment of obtaining and sending a third recommended video is shown, that is, the above step 480, in response to determining that the user portrait information is the same as the initial portrait information, obtaining and sending the third recommended video based on the user portrait information and the initial portrait information may include the following steps:

[0154] Step 610, in response to determining that the user portrait information is the same as the initial portrait information, feature extraction is performed on the user portrait information and the initial portrait information respectively to obtain third feature information and fourth feature information.

[0155] In this step, after the above-mentioned execution entity determines that the user portrait information is the same as the initial portrait information, it can perform feature extraction on the user portrait information and the initial portrait information respectively, and extract meaningful features from the user portrait information and the initial portrait information, which may include the user's behavior, interests, preferences, etc., and extract the same feature information from the user portrait information and the initial portrait information to obtain the third feature information and the fourth feature information.

[0156] Step 620: pre-process the third feature information and the fourth feature information to obtain processed third feature information and processed fourth feature information.

[0157] In this step, the above-mentioned execution entity can preprocess the third characteristic information and the fourth characteristic information, and perform standardization, normalization and other preprocessing on the third characteristic information and the fourth characteristic information to ensure the consistency and comparability of the data, and obtain the processed third characteristic information and the processed fourth characteristic information.

[0158] Step 630: Based on the weight values ​​corresponding to the processed third feature information and the processed fourth feature information, feature fusion is performed on the processed third feature information and the processed fourth feature information to obtain fused feature information.

[0159] In this step, after the above-mentioned execution entity obtains the processed third feature information and the processed fourth feature information, it can assign appropriate weight values ​​to the feature information at each time point in the feature fusion process. The weight value can be determined based on data analysis and business needs to reflect the importance of feature information at different time points.

[0160] Then, the above-mentioned execution entity can perform type analysis on the processed third feature information and the processed fourth feature information. For feature information of the same type, the processed third feature information and the processed fourth feature information can be weighted averaged using a weight value, and the weight value can be adjusted according to business needs; for feature information of different types, the processed third feature information and the processed fourth feature information can be spliced ​​or merged; thereby, the above-mentioned execution entity performs feature fusion on different types of feature information to obtain fused feature information.

[0161] Step 640, based on the fused feature information, obtain the fused portrait information corresponding to the current main user.

[0162] In this step, after the above-mentioned execution entity obtains the fused feature information, it can use a machine learning model to perform feature fusion learning, such as a linear model, decision tree, random forest, etc., input the fused feature information into the learning model, and train the learning model to predict the current main user's behavior or interests, thereby obtaining the fused portrait information corresponding to the current main user based on the fused feature information and the machine learning model.

[0163] Step 650, based on the fused portrait information, obtain and send the third recommended video.

[0164] In this step, after the above-mentioned execution subject obtains the fused portrait information corresponding to the current main user, it can use the fused portrait information to screen all candidate videos, obtain the third recommended video, and send the third recommended video to the client so that the client can play the third recommended video through the screen.

[0165] In this implementation, by fusing the user portrait information and the initial portrait information to obtain the fused portrait information, and using the fused portrait information to obtain and send the third recommended video, the portrait information and the initial portrait information in the temporary shielding process can be compatible at the same time, thereby improving the comprehensiveness of the portrait information, making the obtained recommended videos more comprehensive and accurate, and improving the comprehensiveness and accuracy of the recommended videos.

[0166] Continue to refer Figure 4 , the above video recommendation method may further include the following steps:

[0167] Step 490, in response to determining that the current portrait information is the same as the initial portrait information, obtaining and sending the first video corresponding to the initial portrait information.

[0168] In this step, the execution subject determines that the current portrait information is the same as the initial portrait information, and can use the initial portrait information to obtain the corresponding first video, and transmit the first video to the client so that the client presents the first video through the screen.

[0169] As an example, the execution subject may read the video tag corresponding to the temporary blocking request, preferentially obtain the recommended video corresponding to the video tag, and determine it as the first video.

[0170] Step 4100, based on the operation information of the current main user on the first video, determine the portrait information used to push the video from the current portrait information and the initial portrait information.

[0171] In this step, during the playback of the first video by the client, the operation information of the current main user on the first video can be received, and the operation information can include a temporary blocking operation on the first video, and can also include a viewing operation of watching the first video completely. The above-mentioned execution subject can obtain the operation information of the current main user on the first video, analyze the operation information, determine the viewing status of the current main user on the first video, and determine the portrait information used for pushing the video from the current portrait information and the initial portrait information according to the viewing status.

[0172] As an example, if the operation information is a temporary blocking operation on the first video, and the execution subject determines that the current main user's viewing status of the first video is blocked, the current portrait information continues to be used as the portrait information for video recommendation.

[0173] As an example, if the operation information is a viewing operation of completely watching the first video, the above-mentioned execution subject determines that the current main user's viewing status of the first video is a normal viewing status, and then uses the initial portrait information as the portrait information for video recommendation.

[0174] Step 4200: Obtain and send a third recommended video based on the portrait information.

[0175] In this step, after the above-mentioned execution subject determines the portrait information, it can use the determined portrait information to screen all candidate videos, obtain the third recommended video, and send the third recommended video to the client so that the client can play the third recommended video through the screen.

[0176] In this embodiment, by determining that the current portrait information is the same as the initial portrait information, the first video corresponding to the initial portrait information is obtained, and based on the operation information of the current main user on the first video, the portrait information used to push the video is further determined, so that the original preferred content can be tried to be recommended. If the user continues to choose to block, the blocking is maintained, otherwise the blocking is gradually released and the temporary blocking is lifted. At the same time, the current portrait information and the initial portrait information are used to obtain the third recommended video, thereby realizing the scenario of detecting whether the user has returned to watching the original video and needs to restore the viewing preference, thereby improving the flexibility of video blocking and video recovery.

[0177] refer to Figure 7 , Figure 7 A flowchart 700 of an embodiment of determining the portrait information for pushing a video from the current portrait information and the initial portrait information based on the operation information of the current main user on the first video is shown, that is, the above step 4100, based on the operation information of the current main user on the first video, determining the portrait information for pushing a video from the current portrait information and the initial portrait information, may include the following steps:

[0178] Step 710: In response to determining that the operation information indicates that the first video is playing normally, shorten the push time interval of the first video.

[0179] In this step, the above-mentioned execution subject judges the operation information and determines that the operation information indicates that the first video is played normally. Then, it can be determined that the current main user can watch the recommended video corresponding to the initial portrait information normally, and the push time interval of the first video can be shortened.

[0180] Step 720: Send the second video corresponding to the initial portrait information based on the shortened push time interval.

[0181] In this step, the execution entity may continue to send the second video corresponding to the initial portrait information to the client using the shortened push time interval.

[0182] The above-mentioned execution subject can gradually shorten the push time interval according to the operation information indicating normal playback, and push the video corresponding to the initial portrait information at the shortened time interval.

[0183] Step 730, obtaining the statistical number of operation information indicating normal playback.

[0184] In this step, the execution subject may count the number of times the operation information indicating normal playback is obtained, that is, the number of times the operation information indicating normal playback is obtained during the process of continuously pushing the video corresponding to the initial portrait information.

[0185] Step 720, in response to determining that the statistical number of times exceeds the preset number of times, based on the current portrait information and the initial portrait information, obtain the fused portrait information corresponding to the current main user, and determine the fused portrait information as the portrait information for pushing the video.

[0186] In this step, the above-mentioned execution entity can compare the statistical number of times with the preset number of times. If it is determined that the statistical number of times exceeds the preset number of times, it can be determined that in the process of continuously pushing the video corresponding to the initial portrait information, the number of times the current main user normally watches the video exceeds the preset number of times, and the played video is not operated or watched normally for many times. The current portrait information can be fused with the initial portrait information to obtain the fused portrait information corresponding to the current main user, and the fused portrait information can be determined as the portrait information used to push the video.

[0187] Among them, the above fusion image information can be Figure 6 The steps 610 to 640 are executed to fuse the current portrait information with the initial portrait information to obtain the fused portrait information corresponding to the current main user.

[0188] In this implementation, by continuously pushing the video corresponding to the initial portrait information, the statistical number of times the current main user does not operate or watches normally can be obtained during the continuous pushing process, and when the statistical number exceeds the preset number, the current portrait information is fused with the initial portrait information to obtain the fused portrait information corresponding to the current main user, and the fused portrait information is determined as the portrait information used to push the video. The timing strategy can be adjusted according to various factors such as user operations and interactions. If the user does not operate for many times, the detection time is shortened, making the portrait recovery process more accurate. It is a gradual recovery process, which improves the accuracy of portrait recovery and the flexibility of automatic recovery of portrait information.

[0189] As an optional implementation method, the above step 4100, based on the operation information of the current main user on the first video, determines the portrait information used to push the video from the current portrait information and the initial portrait information, and may also include the following steps: in response to determining that the operation information indicates that the first video is temporarily blocked, execute the video detection step: based on the last push time and push time interval of the first video, determine the current push time of the first video; based on the current push time, send the first video corresponding to the initial portrait information; obtain the operation information of the current main user on the first video; until it is determined that the operation information indicates that the first video is playing normally, determine the portrait information used to push the video from the current portrait information and the initial portrait information.

[0190] Specifically, the above-mentioned execution entity judges the operation information and determines that the operation information indicates that the first video is temporarily blocked. Then, it can be determined that the current main user still needs to temporarily block the first video and continue to maintain the temporarily blocked video restriction. The above-mentioned execution entity can perform the video detection step based on a preset time interval.

[0191] That is, the above-mentioned execution subject can determine the current push time of the first video based on the last push time and the push time interval of the first video. The push time interval can be a pre-set time interval for pushing the video corresponding to the initial portrait information. Then the above-mentioned execution subject can send the first video corresponding to the initial portrait information according to the current push time of the first video, so that the client continues to play the first video. The above-mentioned execution subject can also continue to obtain the operation information of the current main user on the first video. If the operation information still indicates that the first video is temporarily blocked, it is necessary to continue to maintain the temporarily blocked video restriction and continue to execute the video detection step. If the operation information indicates that the first video is playing normally, the portrait information used to push the video can be determined from the current portrait information and the initial portrait information, that is, until the above-mentioned execution subject determines that the operation information indicates that the first video is playing normally, the portrait information used to push the video is determined from the current portrait information and the initial portrait information.

[0192] In this implementation, by continuously pushing the video corresponding to the initial portrait information, during the continuous pushing process, if the operation information indicates that the first video is temporarily blocked, it is necessary to continue to maintain the temporarily blocked video restriction and execute the video detection step, that is, continuously push the first video based on the time interval until it is determined that the operation information indicates that the first video is playing normally, and determine the portrait information used to push the video from the current portrait information and the initial portrait information. The timing strategy can be adjusted according to various factors such as user operations and interactions. If the user continues to choose temporary blocking, the restriction is maintained and the next push time is set. If the user continues to block, the push cycle is extended, so that the portrait recovery process is a gradual recovery process, which improves the accuracy of portrait recovery and the flexibility of automatic recovery of portrait information.

[0193] refer to Figure 8 As an implementation of the methods shown in the above figures, the present disclosure provides an embodiment of a video recommendation device. Figure 2 The method embodiments shown correspond.

[0194] like Figure 8 As shown, the video recommendation device 800 of this embodiment may include: a first acquisition module 810 , a sending module 820 , a second acquisition module 830 , a third acquisition module 840 and a fourth acquisition module 850 .

[0195] The first acquisition module 810 is configured to acquire a first recommended video based on initial portrait information corresponding to the current main user in response to determining the current main user;

[0196] The sending module 820 is configured to send the first recommended video to the client so that the client plays the first recommended video;

[0197] The second acquisition module 830 is configured to, in response to receiving a temporary shielding request for the first recommended video, shield the first recommended video and the associated videos of the first recommended video, and acquire the current portrait information corresponding to the current main user based on the temporary shielding request;

[0198] The third acquisition module 840 is configured to acquire and send the second recommended video based on the current portrait information, so that the client plays the second recommended video;

[0199] The fourth acquisition module 850 is configured to acquire and send a third recommended video based on the current portrait information and the initial portrait information in response to detecting a recovery instruction of the initial portrait information.

[0200] In some optional implementations of the present embodiment, the first acquisition module 810 is further configured to: determine the current main user based on the client's login usage information; obtain the initial portrait information of the current main user based on the historical behavior data of the current main user; and obtain the first recommended video based on the initial portrait information of the current main user.

[0201] In some optional implementations of the present embodiment, the initial portrait information includes a portrait tag and an interest model; and the first acquisition module 810 is further configured to: construct a portrait tag corresponding to the current main user based on the historical behavior data of the current main user; perform feature extraction on the historical behavior data to obtain feature information corresponding to the current main user; based on the portrait tag and the feature information, obtain the interest model corresponding to the current main user; determine the portrait tag and interest model corresponding to the current main user as the initial portrait information of the current main user.

[0202] In some optional implementations of the present embodiment, the first acquisition module 810 is further configured to: perform label screening on the candidate video set based on the portrait label corresponding to the current main user to obtain a first candidate video; perform interest screening on the first candidate video based on the interest model corresponding to the current main user to obtain a second candidate video; and use the interest model to sort the second candidate videos to obtain a first recommended video.

[0203] In some optional implementations of the present embodiment, a temporary blocking request is generated based on the following steps: in response to receiving a temporary blocking operation of the current main user on the first recommended video, obtaining a video tag of the first recommended video; and generating a temporary blocking request based on the temporary blocking operation and the video tag.

[0204] In some optional implementations of the present embodiment, the second acquisition module 830 is further configured to: based on the video tag, add a video filtering condition corresponding to the video tag; based on the video filtering condition, shield the first recommended video and the associated videos of the first recommended video.

[0205] In some optional implementations of the present embodiment, the second acquisition module 830 is further configured to: copy the initial portrait information to obtain copied portrait information; based on the video tag, update the copied portrait information to obtain filtered portrait information; based on the filtered portrait information, shield the first recommended video and the associated videos of the first recommended video.

[0206] In some optional implementations of the present embodiment, the fourth acquisition module 850 is further configured to: determine whether the current portrait information is the same as the initial portrait information; in response to determining that the current portrait information is not the same as the initial portrait information, obtain video playback information within preset conditions; based on the video playback information, obtain user portrait information corresponding to the current main user, and determine whether the user portrait information is the same as the initial portrait information; in response to determining that the user portrait information is the same as the initial portrait information, obtain and send a third recommended video based on the user portrait information and the initial portrait information.

[0207] In some optional implementations of the present embodiment, the fourth acquisition module 850 is further configured to: based on a temporary shielding request, update the current portrait information and the initial portrait information to obtain new current portrait information and new initial portrait information; calculate the portrait similarity between the new current portrait information and the new initial portrait information; based on the portrait similarity, determine whether the current portrait information is the same as the initial portrait information.

[0208] In some optional implementations of the present embodiment, the fourth acquisition module 850 is further configured to: perform feature extraction on the new current portrait information and the new initial portrait information, respectively, to obtain first feature information and second feature information; preprocess the first feature information and the second feature information, to obtain processed first feature information and processed second feature information; calculate the feature similarity between the processed first feature information and the processed second feature information; and calculate the portrait similarity between the new current portrait information and the new initial portrait information based on the feature similarity and the weight value corresponding to the feature similarity.

[0209] In some optional implementations of the present embodiment, the fourth acquisition module 850 is further configured to: in response to determining that the user portrait information is the same as the initial portrait information, respectively extract features from the user portrait information and the initial portrait information to obtain third feature information and fourth feature information; pre-process the third feature information and the fourth feature information to obtain processed third feature information and processed fourth feature information; based on the weight values ​​corresponding to the processed third feature information and the processed fourth feature information, perform feature fusion on the processed third feature information and the processed fourth feature information to obtain fused feature information; based on the fused feature information, obtain fused portrait information corresponding to the current main user; based on the fused portrait information, obtain and send the third recommended video.

[0210] In some optional implementations of the present embodiment, the fourth acquisition module 850 is further configured to: in response to determining that the current portrait information is the same as the initial portrait information, acquire and send the first video corresponding to the initial portrait information; based on the operation information of the current main user on the first video, determine the portrait information used to push the video from the current portrait information and the initial portrait information; based on the portrait information, acquire and send the third recommended video.

[0211] In some optional implementations of the present embodiment, the fourth acquisition module 850 is further configured to: in response to determining that the operation information indicates that the first video is playing normally, shorten the push time interval of the first video; based on the shortened push time interval, send the second video corresponding to the initial portrait information; obtain the statistical number of operation information indicating normal playback; in response to determining that the statistical number exceeds a preset number, based on the user portrait information and the initial portrait information, obtain the fused portrait information corresponding to the current main user, and determine the fused portrait information as the portrait information used to push the video.

[0212] In some optional implementations of the present embodiment, the fourth acquisition module 850 is further configured to: in response to determining that the operation information indicates that the first video is temporarily blocked, execute a video detection step: determine the current push time of the first video based on the last push time and the push time interval of the first video; based on the current push time, send the first video corresponding to the initial portrait information; obtain the operation information of the current main user on the first video; until it is determined that the operation information indicates that the first video is played normally, determine the portrait information used to push the video from the current portrait information and the initial portrait information.

[0213] The video recommendation device provided by the above-mentioned embodiment of the present disclosure, the above-mentioned execution subject first responds to determining the current main user, obtains the first recommended video based on the initial portrait information corresponding to the current main user, and then sends the first recommended video to the client, so that the client plays the first recommended video, and then responds to receiving a temporary shielding request for the first recommended video, shields the first recommended video and the associated video of the first recommended video, and obtains the current portrait information corresponding to the current main user based on the temporary shielding request, and obtains and sends the second recommended video based on the current portrait information, so that the client plays the second recommended video, and finally responds to detecting the recovery instruction of the initial portrait information, obtains and sends the third recommended video based on the current portrait information and the initial portrait information, and then performs During the video recommendation process, a temporary video blocking function can be provided. Based on the temporary blocking request, the first recommended video currently being played and videos of the same type as the first recommended video can be temporarily blocked. The appearance of a certain type of short video can be restricted or the recommendation weight of this type of short video can be reduced within a short period of time. Later, when the instruction to restore the initial portrait information is detected, the third recommended video can be obtained according to the user portrait information and the initial portrait information. That is, after the temporary restriction is lifted, the original short video preference or recommendation weight is restored. The temporary blocking function is introduced to provide users with the ability to temporarily block certain short video categories. After the conditions for restoring the initial portrait information are met, the short video preference recommendation corresponding to the initial portrait information is automatically restored, which improves the flexibility of video recommendation and the targeted video recommendation for different scenarios.

[0214] Those skilled in the art will appreciate that the above device also includes some other well-known structures, such as a processor, a memory, etc. In order to unnecessarily obscure the embodiments of the present disclosure, these well-known structures are described in detail. Figure 8 Not shown.

[0215] It should be noted that the collection, collection, updating, analysis, processing, use, transmission, storage and other aspects of user personal information involved in the technical solution of this disclosure are in compliance with the provisions of relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. Necessary measures are taken for user personal information to prevent illegal access to user personal information data and maintain the security of user personal information, network security and national security.

[0216] Reference below Fig. 9 , which shows a schematic diagram of the structure of an electronic device 900 suitable for implementing the embodiments of the present disclosure. The terminal device in the embodiments of the present disclosure may include but is not limited to mobile terminals such as smart screens, notebook computers, PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Fig. 9 The terminal device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0217] like Fig. 9 As shown, the electronic device 900 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage device 908 into a random access memory (RAM) 903. In the RAM 903, various programs and data required for the operation of the electronic device 900 are also stored. The processing device 901, the ROM 902, and the RAM 903 are connected to each other via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0218] Typically, the following devices may be connected to the I / O interface 905: an input device 906 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 907 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 908 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 909. The communication device 909 may allow the electronic device 900 to communicate with other devices wirelessly or by wire to exchange data. Although Fig. 9 The electronic device 900 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead. Fig. 9 Each block shown in the figure may represent one device, or may represent multiple devices as required.

[0219] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through a communication device 909, or installed from a storage device 908, or installed from a ROM 902. When the computer program is executed by the processing device 901, the above functions defined in the method of the embodiment of the present disclosure are executed. It should be noted that the computer-readable medium of the embodiment of the present disclosure can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection with one or more conductors, a portable computer disk, 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In an embodiment of the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, an apparatus, or a device. In an embodiment of the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in combination with an instruction execution system, an apparatus, or a device. The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to: wire, optical cable, RF (radio frequency), etc., or any suitable combination of the foregoing.

[0220] Computer program code for performing the operations of the embodiments of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet).

[0221] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0222] The units involved in the embodiments described in the present application may be implemented by software or hardware. The units described may also be set in a processor, for example, it may be described as: a processor includes a first acquisition module, a sending module, a second acquisition module, a third acquisition module and a fourth acquisition module, wherein the names of these modules do not constitute a limitation on the modules themselves in some cases.

[0223] As another aspect, the present application also provides a computer-readable medium, which may be included in the electronic device; or may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs, and when the one or more programs are executed by the electronic device, the electronic device: in response to determining the current main user, based on the initial portrait information corresponding to the current main user, obtains a first recommended video; sends the first recommended video to the client so that the client plays the first recommended video; in response to receiving a temporary shielding request for the first recommended video, shields the first recommended video and the associated video of the first recommended video, and obtains the current portrait information corresponding to the current main user based on the temporary shielding request; based on the current portrait information, obtains and sends the second recommended video so that the client plays the second recommended video; in response to detecting an instruction to restore the initial portrait information, obtains and sends a third recommended video based on the current portrait information and the initial portrait information.

[0224] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by a specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with the technical features with similar functions disclosed in the embodiments of the present disclosure (but not limited to) to form a technical solution.

Claims

1. A video recommendation method, the method comprising: In response to determining the current main user, obtaining a first recommended video based on initial portrait information corresponding to the current main user; Sending the first recommended video to a client, so that the client plays the first recommended video; In response to receiving a temporary shielding request for the first recommended video, shielding the first recommended video and the associated videos of the first recommended video, and acquiring current portrait information corresponding to the current main user based on the temporary shielding request; Based on the current portrait information, obtaining and sending a second recommended video, so that the client plays the second recommended video; In response to detecting a recovery instruction for the initial portrait information, a third recommended video is acquired and sent based on the current portrait information and the initial portrait information.

2. The method according to claim 1, wherein: In response to determining the current main user, obtaining a first recommended video based on initial portrait information corresponding to the current main user includes: Determine the current primary user based on the login usage information of the client; Based on the historical behavior data of the current main user, obtaining initial portrait information of the current main user; Based on the initial portrait information of the current main user, a first recommended video is obtained.

3. The method according to claim 2, wherein: The initial portrait information includes a portrait tag and an interest model; and the acquiring the initial portrait information of the current primary user based on the historical behavior data of the current primary user includes: Based on the historical behavior data of the current main user, construct a portrait label corresponding to the current main user; Extracting features from the historical behavior data to obtain feature information corresponding to the current primary user; Based on the portrait tag and the feature information, acquiring an interest model corresponding to the current primary user; The portrait label and interest model corresponding to the current main user are determined as the initial portrait information of the current main user.

4. The method according to claim 3, wherein: The acquiring a first recommended video based on the initial portrait information of the current primary user includes: Based on the portrait tag corresponding to the current main user, the candidate video set is subjected to tag screening to obtain a first candidate video; Based on the interest model corresponding to the current primary user, the first candidate video is screened for interest to obtain a second candidate video; The second candidate videos are sorted by using the interest model to obtain the first recommended video.

5. The method according to claim 1, wherein: The temporary blocking request is generated based on the following steps: In response to receiving a temporary blocking operation of the current primary user on the first recommended video, acquiring a video tag of the first recommended video; The temporary blocking request is generated based on the temporary blocking operation and the video tag.

6. The method according to claim 5, wherein: The shielding of the first recommended video and the associated videos of the first recommended video includes: Based on the video tag, adding a video filtering condition corresponding to the video tag; Based on the video filtering condition, the first recommended video and the associated videos of the first recommended video are shielded.

7. The method according to claim 5, wherein: The shielding of the first recommended video and the associated videos of the first recommended video includes: Copying the initial portrait information to obtain copied portrait information; Based on the video tag, the copied portrait information is updated to obtain filtered portrait information; Based on the filtering portrait information, the first recommended video and the associated videos of the first recommended video are shielded.

8. The method according to claim 1, wherein in response to detecting the restoration instruction of the initial portrait information, acquiring and sending a third recommended video based on the current portrait information and the initial portrait information comprises: Determining whether the current portrait information is the same as the initial portrait information; In response to determining that the current portrait information is different from the initial portrait information, obtaining video playback information within a preset condition; Based on the video playback information, obtaining user portrait information corresponding to the current primary user, and determining whether the user portrait information is the same as the initial portrait information; In response to determining that the user portrait information is the same as the initial portrait information, a third recommended video is obtained and sent based on the user portrait information and the initial portrait information.

9. The method according to claim 8, wherein: The determining whether the current portrait information is the same as the initial portrait information includes: Based on the temporary shielding request, the current portrait information and the initial portrait information are updated to obtain new current portrait information and new initial portrait information; Calculating the portrait similarity between the new current portrait information and the new initial portrait information; Based on the portrait similarity, it is determined whether the current portrait information is the same as the initial portrait information.

10. The method according to claim 9, wherein: The calculating the portrait similarity between the new current portrait information and the new initial portrait information includes: Extracting features from the new current portrait information and the new initial portrait information respectively to obtain first feature information and second feature information; Preprocessing the first feature information and the second feature information to obtain processed first feature information and processed second feature information; Calculating feature similarity between the processed first feature information and the processed second feature information; Based on the feature similarity and the weight value corresponding to the feature similarity, the portrait similarity between the new current portrait information and the new initial portrait information is calculated.

11. The method according to claim 8, wherein: In response to determining that the user portrait information is the same as the initial portrait information, acquiring and sending a third recommended video based on the user portrait information and the initial portrait information includes: In response to determining that the user portrait information is the same as the initial portrait information, extracting features from the user portrait information and the initial portrait information to obtain third feature information and fourth feature information; Preprocessing the third characteristic information and the fourth characteristic information to obtain processed third characteristic information and processed fourth characteristic information; Based on the weight values ​​corresponding to the processed third feature information and the processed fourth feature information, performing feature fusion on the processed third feature information and the processed fourth feature information to obtain fused feature information; Based on the fused feature information, obtaining fused portrait information corresponding to the current primary user; Based on the fused portrait information, a third recommended video is obtained and sent.

12. The method according to any one of claims 8 or 11, wherein: The step of acquiring and sending a third recommended video based on the current portrait information and the initial portrait information in response to detecting a recovery instruction of the initial portrait information further includes: In response to determining that the current portrait information is the same as the initial portrait information, acquiring and sending a first video corresponding to the initial portrait information; Based on the operation information of the current main user on the first video, determining the portrait information for pushing the video from the current portrait information and the initial portrait information; Based on the portrait information used to push the video, a third recommended video is obtained and sent.

13. The method according to claim 12, wherein: The determining, based on the operation information of the current main user on the first video, the portrait information for pushing the video from the current portrait information and the initial portrait information includes: In response to determining that the operation information indicates that the first video is played normally, shortening the push time interval of the first video; Based on the shortened push time interval, sending a second video corresponding to the initial portrait information; Obtaining a statistical number of the operation information indicating normal playback; In response to determining that the statistical number of times exceeds a preset number of times, based on the current portrait information and the initial portrait information, the fused portrait information corresponding to the current main user is obtained, and the fused portrait information is determined as the portrait information for pushing the video.

14. The method according to claim 13, wherein: The determining, based on the operation information of the current main user on the first video, the portrait information for pushing the video from the current portrait information and the initial portrait information further includes: In response to determining that the operation information indicates that the first video is temporarily blocked, performing a video detection step: determining a current push time of the first video based on a last push time and a push time interval of the first video; sending the first video corresponding to the initial portrait information based on the current push time; and obtaining operation information of the current primary user on the first video; Until it is determined that the operation information indicates that the first video is played normally, the portrait information used to push the video is determined from the current portrait information and the initial portrait information.

15. A video recommendation device, comprising: A first acquisition module is configured to, in response to determining a current main user, acquire a first recommended video based on initial portrait information corresponding to the current main user; A sending module, configured to send the first recommended video to a client, so that the client plays the first recommended video; a second acquisition module configured to, in response to receiving a temporary shielding request for the first recommended video, shield the first recommended video and the associated videos of the first recommended video, and acquire current portrait information corresponding to the current main user based on the temporary shielding request; A third acquisition module is configured to acquire and send a second recommended video based on the current portrait information, so that the client plays the second recommended video; The fourth acquisition module is configured to acquire and send a third recommended video based on the current portrait information and the initial portrait information in response to detecting a recovery instruction of the initial portrait information.

16. An electronic device, comprising: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors are enabled to implement the method according to any one of claims 1 to 14.

17. A computer readable medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 14 is implemented.

Images