Video recall method and device, storage medium and electronic equipment

Abstract

The invention relates to a video recall method and device, a storage medium and electronic equipment. The method comprises the following steps: when a video uploading request is received, acquiring a first video requested to be uploaded by the video uploading request and a video tag of the first video; when the first video comprises an associated video, constructing a label relation tree according to a video label of the first video and a video label of the associated video; when a video search request of a target user is received, determining a searched target video according to the video search request; under the condition that the video tag of the target video is located in the tag relation tree, obtaining all videos corresponding to all tags in the tag relation tree; and pushing the video identifiers of all videos to the target user. The technical problems that a series of intentions of a user cannot be understood and videos in all related series cannot be effectively recalled are solved.

Description

Technical Field

[0001] The present application relates to the field of video recall, and in particular to a video recall method, device, storage medium and electronic device. Background Art

[0002] In the prior art, users can search for video works through video platforms, such as searching for multimedia programs on TV or on mobile devices for viewing. However, when searching for video works, it is often impossible to effectively recall all video works in related series. This is because traditional search engines do not have the ability to understand users' complex query intentions, especially for queries with series intentions, and the search results are often unsatisfactory, which greatly reduces the accuracy of the search and the user experience. Summary of the invention

[0003] The present application provides a video recall method, device, storage medium and electronic device to solve the technical problem of being unable to understand the user's series intention and being unable to effectively recall all videos in the relevant series.

[0004] In a first aspect, the present application provides a video recall method, comprising: upon receiving a video upload request, obtaining a first video requested to be uploaded by the video upload request and a video tag of the first video; when the first video includes associated associated videos, constructing a tag relationship tree according to the video tag of the first video and the video tag of the associated videos; upon receiving a video search request from a target user, determining a searched target video according to the video search request; when the video tag of the target video is located in the tag relationship tree, obtaining all videos corresponding to all tags in the tag relationship tree; and pushing the video identifiers of all the videos to the target user.

[0005] In a second aspect, the present application provides a video recall device, comprising: an acquisition module, for acquiring, upon receiving a video upload request, a first video requested to be uploaded by the video upload request and a video tag of the first video; a construction module, for constructing a tag relationship tree according to the video tag of the first video and the video tag of the associated video when the first video includes associated associated videos; a determination module, for determining, upon receiving a video search request from a target user, a searched target video according to the video search request; an acquisition module, for acquiring, when the video tag of the target video is located in the tag relationship tree, all videos corresponding to all tags in the tag relationship tree; a push module, for pushing the video identifiers of all the videos to the target user.

[0006] As an optional example, the acquisition module includes: a generation unit, used to obtain the video classification of the first video from the video information of the first video; based on the video classification, generate the video tag for the first video, wherein the video tag is used to uniquely identify the first video, and the video tag indicates the video classification of the first video.

[0007] As an optional example, the above-mentioned construction module includes: a construction unit, used to determine the association relationship between the above-mentioned first video and the above-mentioned associated video, wherein the above-mentioned association relationship includes a parallel relationship and a parent-child relationship; when the above-mentioned association relationship is a parallel relationship, a parallel label relationship tree is generated; when the above-mentioned association relationship is a parent-child relationship, a parent-child label relationship tree is generated.

[0008] As an optional example, the construction unit includes: a determination subunit, used to determine the association relationship with the associated video through the associated tag in the video information of the first video; or, to obtain the title information in the video information of the first video, and to determine, from multiple videos according to the title information, a video with a title that is highly similar to the title information and has a serial number, and to use the determined video as the first video and determine the association relationship.

[0009] As an optional example, the above-mentioned determination module includes: a determination unit, used to determine the user intention based on the keywords in the above-mentioned video search request; determine a video that is the same or similar to the above-mentioned user intention from the video information of the video; and use the determined video as the above-mentioned target video.

[0010] As an optional example, the acquisition module includes: a traversal unit, which is used to, when the video tag of the target video is located in the tag relationship tree, start from the video tag of the target video, traverse the parent node and the parent node of each parent node upward until the root node is traversed; start from the root node, traverse each child node and the child node of each child node until all child nodes that do not contain any child nodes are traversed; and use the video corresponding to the tag of each traversed node as one of the above-mentioned videos.

[0011] As an optional example, the push module includes: a push unit, used to sort all the videos according to the video information of each video in all the videos; and push the video identifiers of all the sorted videos to the target user.

[0012] In a third aspect, the present application provides an electronic device comprising: at least one communication interface; at least one bus connected to the at least one communication interface; at least one processor connected to the at least one bus; and at least one memory connected to the at least one bus, wherein the memory stores a computer program, and the processor is configured to implement any one of the above-mentioned video recall methods when executing the computer program.

[0013] In a fourth aspect, the present application further provides a computer storage medium storing computer executable instructions, wherein the computer executable instructions are used to execute any of the above-mentioned video recall methods of the present application.

[0014] The above-mentioned technical scheme provided by the embodiment of the present application has the following advantages over the prior art: the scheme provided by the embodiment of the present application, when receiving a video upload request, obtains the first video requested to be uploaded by the above-mentioned video upload request and the video tag of the above-mentioned first video; when the above-mentioned first video includes associated associated videos, constructs a tag relationship tree according to the video tag of the above-mentioned first video and the video tag of the above-mentioned associated videos; when receiving a video search request of a target user, determines the searched target video according to the above-mentioned video search request; when the video tag of the above-mentioned target video is located in the above-mentioned tag relationship tree, obtains all videos corresponding to all tags in the above-mentioned tag relationship tree; pushes the video identifiers of the above-mentioned all-mentioned videos to the above-mentioned target user, so that when the target user performs a video search, the target video that meets the user's search intention and the video series to which the target video belongs can be pushed to the target user, so as to understand the user's series intention and effectively recall videos under all related series. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0017] One or more embodiments are exemplarily described by pictures in the corresponding drawings, and these exemplified descriptions do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings represent similar elements, and unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0018] Figure 1A flowchart of a video recall method provided in an embodiment of the present application;

[0019] Figure 2 A label relationship tree of a video recall method provided in an embodiment of the present application;

[0020] Figure 3 A schematic diagram of a traversal label relationship tree of a video recall method provided in an embodiment of the present application;

[0021] Figure 4 A label relationship tree sorting diagram of a video recall method provided in an embodiment of the present application;

[0022] Figure 5 A schematic diagram of the structure of a video recall device provided in an embodiment of the present application;

[0023] Figure 6 A schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0025] The disclosure below provides many different embodiments or examples to implement different structures of the present invention. In order to simplify the disclosure of the present invention, the parts and settings of specific examples are described below. Of course, they are only examples, and the purpose is not to limit the present invention. In addition, the present invention can repeat reference numbers and / or letters in different examples. This repetition is for the purpose of simplification and clarity, and does not itself indicate the relationship between the various embodiments and / or settings discussed.

[0026] In order to solve the technical problem that the prior art cannot understand the user's series intention and cannot effectively recall all videos in the related series, the present application provides a video recall method that can achieve the purpose of understanding the user's series intention and effectively recalling all videos in the related series.

[0027] Figure 1 A flowchart of a video recall method provided in an embodiment of the present application. Figure 1 As shown, the above video recall method includes:

[0028] S102, upon receiving a video upload request, obtaining a first video requested to be uploaded by the video upload request and a video tag of the first video;

[0029] S104, when the first video includes associated associated videos, constructing a tag relationship tree according to the video tag of the first video and the video tags of the associated videos;

[0030] S106, upon receiving a video search request from a target user, determining a searched target video according to the video search request;

[0031] S108, when the video tag of the target video is in the tag relationship tree, obtaining all videos corresponding to all tags in the tag relationship tree;

[0032] S110: Push video identifiers of all videos to target users.

[0033] The present application can be applied on the server side of a video platform, where the server side responds to a video upload request, stores the first uploaded video on the server side, and responds to a target user's video search request, pushes the videos stored on the server side that meet the requirements to the target user, and the target user selects a video to play on the client's video platform.

[0034] For example, the above method is applied to the process of video-on-demand multimedia programs. As a platform for TV programs, such as a certain audio and video software, the content of the 7th episode of the 3rd season of a certain children's program is uploaded. Then, the server can respond to the video upload request and store the content of the 7th episode of the 3rd season of the children's program on the server side. In addition, according to the content of the 7th episode of the 3rd season of the children's program, it will search whether the server has uploaded the first 6 episodes of the 3rd season, whether there are the content of the last few episodes of the 3rd season, and whether there are the content of the first season, the second season, the fourth season, the fifth season..., that is, check whether there are related videos on the server side. If there are related videos, the number of tags is constructed and the association relationship is recorded. When the user is watching TV, for example, searching for a certain episode of the children's program, or searching for the children's program, the server will return the entire series of the third season of the children's program searched by the user (the identification of each episode, to avoid returning all videos and occupying network resources, if there are the content of the first season, the second season, the fourth season, the fifth season..., they will all be returned) to the user for selection.

[0035] The target user in this embodiment may be a user watching TV. When uploading a video, the engineer of the audio and video software may upload the video and store it in the server.

[0036] In this embodiment, when uploading the first video, the server obtains the first video and the video tag of the first video, and determines whether the first video includes associated videos in the server through the video tags. If the first video does not include associated videos, the first video can be stored. If the first video includes associated videos in the server, it means that the associated videos of the first video have been uploaded previously, and the video tag of the first video is to construct a tag relationship tree with the video tag of the associated video, so that the relationship between the first video and the associated videos of the first video is established through the tags, and all other videos can be found through any one and the tag relationship tree.

[0037] The above video tags can uniquely mark the video. The video tags can be generated according to certain rules. For example, the video tags can be generated according to the upload time of the video + the video title, or can be uniquely assigned by the server.

[0038] In this embodiment, the relationship between the first video and the associated video can be a parent-child relationship or a parallel relationship, or an indirect parent-child relationship or an indirect parallel relationship. For example, the first video can be parallel to the associated video, and when there are multiple associated videos, the associated videos can also be in a parent-child relationship or a parallel relationship. For example, the associated videos include associated videos 1 to 3, the first video is parallel to associated video 1, associated video 1 is the parent video of associated video 2, and associated video 2 is parallel to associated video 3.

[0039] After the server constructs a tag relationship library using video tags, the server side retains the relationship between the first video and the associated videos. Figure 2 These are multiple exemplary label relationship diagrams, each of which marks a relationship between a first video and at least one associated video of the first video. Figure 2 In the three relationship diagrams, arrows point from parent nodes to child nodes, and each node is a video label.

[0040] If a video search request from a target user is received, the video that the user wants to watch can be searched according to the video search request. After searching for the target video, determine whether the video tag of the target video exists in the above tag relationship graph. If the video tag of the target video appears in any tag relationship graph, traverse the tag relationship graph to obtain all tags of the tag relationship graph, and push the videos corresponding to all tags as a collection to the target user. If multiple tag relationship graphs are found according to the video tag of the target video, multiple collections are obtained. If no tag relationship graph is found, the target video is pushed to the target user. There can be one or more target videos, and the above tag relationship graph must be searched for each target video.

[0041] For example, Figure 3As shown, for the target video, when the video tag is found on the first tag relationship graph, traverse to the root node first to node 1, and then traverse downward from the root node to nodes 2-7 (the video tag of the target video is node 7, not shown in the figure), and obtain the video collection where the target video is located, that is, the collection of videos corresponding to video tag nodes 1-7.

[0042] The solution provided by the embodiment of the present application, when receiving a video upload request, obtains the first video requested to be uploaded by the video upload request and the video tag of the first video; when the first video includes associated associated videos, builds a tag relationship tree according to the video tag of the first video and the video tag of the associated videos; when receiving a video search request from a target user, determines the searched target video according to the video search request; when the video tag of the target video is located in the tag relationship tree, obtains all videos corresponding to all tags in the tag relationship tree; pushes the video identifiers of all videos to the target user, so that when the target user performs a video search, the target video that meets the user's search intention and the video series to which the target video belongs can be pushed to the target user, thereby achieving the effect of understanding the user's series intention and effectively recalling videos under all related series.

[0043] As an optional example, obtaining the video tag of the first video includes: obtaining the video classification of the first video from the video information of the first video; generating a video tag for the first video according to the video classification, wherein the video tag is used to uniquely identify the first video, and the video tag indicates the video classification of the first video.

[0044] In this embodiment, the video tag of the first video may be input or set by the user, or may be set by the system. If it is input or set by the user, the video tag may carry the keywords input by the user, and may be marked with associated videos associated with the first video. Through video tags, users can construct different videos into videos that are associated with each other, thereby forming a video collection of multiple videos. For example, after uploading video 1, when uploading video 2, the user adds a video tag to video 2, thereby making video 1 an associated video of video 2, and forming a video collection with video 1 and video 2, and constructing a tag relationship tree.

[0045] If the system generates video tags, the system can identify the titles of the videos, classify the videos with high title similarity as videos in a collection, and add video tags of the category to them. In other words, the system can compare the first video uploaded by the uploader with other videos stored in the server, compare the similarity of the titles, and check whether the first video has previously uploaded related videos.

[0046] As an optional example, when the first video includes associated associated videos, constructing a tag relationship tree based on the video tags of the first video and the video tags of the associated videos includes: determining the association relationship between the first video and the associated videos, wherein the association relationship includes a parallel relationship and a parent-child relationship; when the association relationship is a parallel relationship, generating a parallel tag relationship tree; when the association relationship is a parent-child relationship, generating a parent-child tag relationship tree.

[0047] In this embodiment, when the first video contains an associated video, it has been determined that the first video and the associated video need to construct a tag relationship tree. At this time, it is necessary to determine the association relationship between the first video and the associated video, whether it is a parent-child relationship or a parallel relationship. Among them, the parent-child relationship or the parallel relationship does not refer to which of the two videos contains which one, or the two are parallel, but the display order of the first video and the associated video in the collection is determined by the parent-child relationship and the parallel relationship. For the parent-child relationship, for example, a parent video is followed by two child videos, and the playback order of the two child videos is after the parent video. If the two child videos are in a parallel relationship, the playback order can be reversed.

[0048] As an optional example, determining the association relationship between the first video and the associated video includes: determining the association relationship with the associated video through the association tag in the video information of the first video; or, obtaining the title information in the video information of the first video, and determining, from multiple videos based on the title information, a video with a high degree of similarity between the title and the title information and with a serial number, taking the determined video as the first video and determining the association relationship.

[0049] In this embodiment, when determining the association relationship between the first video and the associated videos, if the user determines the video tag, the associated videos related to the first video have been determined through the video tag of the first video, and the association relationship between the first video and the associated videos is marked as a parent-child relationship or a parallel relationship. If the system identifies the association relationship, the system can identify the first video and multiple videos, recognize that the title of the first video is highly similar to the titles of the multiple videos, and the first video and the multiple videos are numbered. At this time, it can be determined that the first video is associated with the multiple videos, and the association relationship is determined based on the serial number. The parent video is in the first serial number, the child video is in the second serial number, and the parallel relationship is the same for the serial numbers.

[0050] As an optional example, when a video search request from a target user is received, determining the searched target video based on the video search request includes: determining the user intent based on keywords in the video search request; determining a video that is the same or similar to the user intent from the video information of the video; and using the determined video as the target video.

[0051] In this embodiment, when the target user initiates a video search request, keywords can be extracted from the video search request, and then the user intent matching the keyword can be found based on the keyword. This process can be determined based on the neural network model. By inputting keywords, the neural network model gives the user intent, and analyzes the user intent to obtain the video that the user may want to search for. Among the results output by the model, the video with a higher possibility is used as the target video.

[0052] As an optional example, when the video tag of the target video is located in a tag relationship tree, obtaining all videos corresponding to all tags in the tag relationship tree includes: when the video tag of the target video is located in the tag relationship tree, starting from the video tag of the target video, traversing upward through the parent nodes and the parent nodes of each parent node until the root node is traversed; starting from the root node, traversing each child node and the child nodes of each child node until all child nodes that do not contain any child nodes are traversed; and taking the video corresponding to the tag of each traversed node as one of all videos.

[0053] In this embodiment, after acquiring the target video, the video tag of the target video is determined, and then the tag is searched from all tag relationship trees. After the relationship tree of the tag is found, traversal is performed to traverse all tags in the relationship tree.

[0054] The traversal method starts from the tag of the target video found, traverses upward to the root node, and then traverses downward from the root node to each node that does not contain child nodes. Thus, all nodes in each tag relationship tree are traversed. Each tag relationship tree traverses to obtain multiple video tags, and the multiple videos corresponding to the multiple video tags under the tag relationship tree constitute a video collection. Multiple tag relationship trees constitute multiple video collections.

[0055] In this embodiment, when the target user wants to search for videos and initiates a video search request, the server searches for multiple target videos based on the video search request, but does not recommend the target videos to the target user. Instead, the server recommends a video collection containing each target video among the multiple target videos to the target user.

[0056] As an optional example, pushing the video identifiers of all videos to the target user includes: sorting all videos according to the video information of each video in all videos; and pushing the video identifiers of all sorted videos to the target user.

[0057] In this embodiment, when each video collection is recommended to the user, the videos in each video collection must be sorted and then recommended to the user after sorting.

[0058] There are many ways to sort, and the sorting order can be determined according to the parent-child relationship or parallel relationship in the tag relationship tree. One sorting method is to start from the root node, with the root node as the first video, and then traverse the child nodes of the root node. The videos of the child nodes are located after the root node. If there are multiple child nodes, the order of the multiple child nodes is parallel, and they can be sorted randomly or according to the order of video upload time. The child nodes continue to be traversed until the last node is traversed.

[0059] For example, Figure 4 As shown, Figure 4 This is a schematic diagram of different sorting results for the same tag relationship tree in this embodiment. For root node 1, there is only one, and then traverse the child nodes. If there are multiple child nodes, they are in a parallel relationship. The parallel relationship is randomly sorted. For example, the first and second cases are sorted as sort 2 and sort 5, and the third and fourth cases are sorted as sort 3 and sort 2. For non-parallel cases, such as sort 4 in the first case, there is only one child node, so it is directly sorted after sort 4, which is sort 5. After sorting, 6 videos are composed of Figure 4 There are four sorting situations in .

[0060] Another sorting method is to perform comprehensive sorting based on the relevance between the keywords provided by the user during the search and the recalled videos (from high to low), the upload time of the recalled videos (from recent to distant), and other sorting strategies (the purpose is to put the videos with the strongest relevance to the user's search intention in the first place, while also displaying other videos in the series to the user at the same time to meet the user's search needs and search experience).

[0061] This application first uses the video tags mounted on the video, for example, the tags can be video series identifiers (Identity document, id), to establish a parent-child series relationship tree (parent-child tag relationship tree) and a peer series relationship tree (parallel tag relationship tree); then, for the search request input by the target user, the series intention is determined and the corresponding series id is obtained; then, the parent-child series relationship tree and the peer series relationship tree are traversed and queried through the series id to obtain the corresponding main series id (root node id); then, all videos corresponding to the series id under the parent-child series relationship tree or the peer series relationship tree are recalled according to the main series id; finally, all recalled videos are sorted on the basis of full recall. After sorting, they are pushed to users.

[0062] Constructing a series tree: Using the series identification code (series ID) on the video document, establish a parent-child series relationship tree and a peer series relationship tree.

[0063] Identify query intent: Analyze and determine the query request entered by the user, find out its series intent, and obtain its corresponding series ID;

[0064] Traverse the series tree: Use the series ID to traverse the series relationship tree and find the main series ID;

[0065] Video recall: Based on the main series ID, all related series videos are recalled together;

[0066] Result sorting: Sort the recalled videos so that users can find the videos they need more conveniently and quickly.

[0067] When uploading each video, we check whether the server has stored related videos of the video. If there are related videos, we build a tag relationship tree. If there are no related videos, the uploaded video can form a series. After building the tag relationship tree, if the user searches for any video while watching TV, the entire series will be recommended to the user.

[0068] For example, when a user uploads the second episode of the third season of a children's program, it checks whether the content of other seasons of the children's program and other episodes of the third season are stored. If they are stored, a label relationship tree is established. Programs of different seasons are in a parallel relationship and are located under the same parent node. Different episodes of the same season are in a parallel relationship and are located under the same season point. If the user's search keyword hits any episode of any season of the children's program, the label relationship tree is traversed to find the content of each episode of each season of the children's program stored on the server, and then sent to the user (sending identification) after sorting for the user to choose.

[0069] This embodiment organizes and indexes the video content under the series through a series tree structure. The series tree structure can effectively reflect the series relationship between videos, including parent-child series relationship and peer series relationship. This structure can make the recall of all video content very efficient. Even if the user's query content is slightly biased, this application takes all the video content under the relevant series into consideration through the series tree traversal, thereby greatly enhancing the accuracy and comprehensiveness of the search. It improves the problems of low accuracy of search results and incomplete search scope that are common in current network searches.

[0070] Figure 5 The structure diagram of a video recall device provided in an embodiment of the present application is shown in FIG. Figure 5 As shown, the video recall device comprises:

[0071] A first acquisition module 502 is used to acquire, upon receiving a video upload request, a first video requested to be uploaded by the video upload request and a video tag of the first video;

[0072] A construction module 504 is used to construct a tag relationship tree according to the video tag of the first video and the video tags of the associated videos when the first video includes associated associated videos;

[0073] A determination module 506, configured to determine the searched target video according to the video search request when receiving a video search request from a target user;

[0074] The second acquisition module 508 is used to acquire all videos corresponding to all tags in the tag relationship tree when the video tag of the target video is located in the tag relationship tree;

[0075] The push module 510 is used to push the video identifiers of all videos to target users.

[0076] The present application can be applied on the server side of a video platform, where the server side responds to a video upload request, stores the first uploaded video on the server side, and responds to a target user's video search request, pushes the videos stored on the server side that meet the requirements to the target user, and the target user selects a video to play on the client's video platform.

[0077] For example, the above method is applied to the process of video-on-demand multimedia programs. As a platform for TV programs, such as a certain audio and video software, the content of the 7th episode of the 3rd season of a certain children's program is uploaded. Then, the server can respond to the video upload request and store the content of the 7th episode of the 3rd season of the children's program on the server side. In addition, according to the content of the 7th episode of the 3rd season of the children's program, it will search whether the server has uploaded the first 6 episodes of the 3rd season, whether there are the content of the last few episodes of the 3rd season, and whether there are the content of the first season, the second season, the fourth season, the fifth season..., that is, check whether there are related videos on the server side. If there are related videos, the number of tags is constructed and the association relationship is recorded. When the user is watching TV, for example, searching for a certain episode of the children's program, or searching for the children's program, the server will return the entire series of the third season of the children's program searched by the user (the identification of each episode, to avoid returning all videos and occupying network resources, if there are the content of the first season, the second season, the fourth season, the fifth season..., they will all be returned) to the user for selection.

[0078] The target user in this embodiment may be a user watching TV. When uploading a video, the engineer of the audio and video software may upload the video and store it in the server.

[0079] In this embodiment, when uploading the first video, the server obtains the first video and the video tag of the first video, and determines whether the first video includes associated videos in the server through the video tags. If the first video does not include associated videos, the first video can be stored. If the first video includes associated videos in the server, it means that the associated videos of the first video have been uploaded previously, and the video tag of the first video is to construct a tag relationship tree with the video tag of the associated video, so that the relationship between the first video and the associated videos of the first video is established through the tags, and all other videos can be found through any one and the tag relationship tree.

[0080] The above video tags can uniquely mark the video. The video tags can be generated according to certain rules. For example, the video tags can be generated according to the upload time of the video + the video title, or can be uniquely assigned by the server.

[0081] In this embodiment, the relationship between the first video and the associated video can be a parent-child relationship or a parallel relationship, or an indirect parent-child relationship or an indirect parallel relationship. For example, the first video can be parallel to the associated video, and when there are multiple associated videos, the associated videos can also be in a parent-child relationship or a parallel relationship. For example, the associated videos include associated videos 1 to 3, the first video is parallel to associated video 1, associated video 1 is the parent video of associated video 2, and associated video 2 is parallel to associated video 3.

[0082] After the server constructs a tag relationship library using video tags, the server side retains the relationship between the first video and the associated videos. Figure 2 These are multiple exemplary label relationship diagrams, each of which marks a relationship between a first video and at least one associated video of the first video. Figure 2 In the three relationship diagrams, arrows point from parent nodes to child nodes, and each node is a video label.

[0083] If a video search request from a target user is received, the video that the user wants to watch can be searched according to the video search request. After searching for the target video, determine whether the video tag of the target video exists in the above tag relationship graph. If the video tag of the target video appears in any tag relationship graph, traverse the tag relationship graph to obtain all tags of the tag relationship graph, and push the videos corresponding to all tags as a collection to the target user. If multiple tag relationship graphs are found according to the video tag of the target video, multiple collections are obtained. If no tag relationship graph is found, the target video is pushed to the target user. There can be one or more target videos, and the above tag relationship graph must be searched for each target video.

[0084] For example, Figure 3As shown, for the target video, when the video tag is found on the first tag relationship graph, traverse to the root node first to node 1, and then traverse downward from the root node to nodes 2-7 (the video tag of the target video is node 7, not shown in the figure), and obtain the video collection where the target video is located, that is, the collection of videos corresponding to video tag nodes 1-7.

[0085] The solution provided by the embodiment of the present application, when receiving a video upload request, obtains the first video requested to be uploaded by the video upload request and the video tag of the first video; when the first video includes associated associated videos, builds a tag relationship tree according to the video tag of the first video and the video tag of the associated videos; when receiving a video search request from a target user, determines the searched target video according to the video search request; when the video tag of the target video is located in the tag relationship tree, obtains all videos corresponding to all tags in the tag relationship tree; pushes the video identifiers of all videos to the target user, so that when the target user performs a video search, the target video that meets the user's search intention and the video series to which the target video belongs can be pushed to the target user, thereby achieving the effect of understanding the user's series intention and effectively recalling videos under all related series.

[0086] As an optional example, the above-mentioned first acquisition module includes: a generation unit, used to obtain the video classification of the first video from the video information of the first video; based on the video classification, generate a video tag for the first video, wherein the video tag is used to uniquely identify the first video, and the video tag indicates the video classification of the first video.

[0087] In this embodiment, the video tag of the first video may be input or set by the user, or may be set by the system. If it is input or set by the user, the video tag may carry the keywords input by the user, and may be marked with associated videos associated with the first video. Through video tags, users can construct different videos into videos that are associated with each other, thereby forming a video collection of multiple videos. For example, after uploading video 1, when uploading video 2, the user adds a video tag to video 2, thereby making video 1 an associated video of video 2, and forming a video collection with video 1 and video 2, and constructing a tag relationship tree.

[0088] If the system generates video tags, the system can identify the titles of the videos, classify the videos with high title similarity as videos in a collection, and add video tags of the category to them. In other words, the system can compare the first video uploaded by the uploader with other videos stored in the server, compare the similarity of the titles, and check whether the first video has previously uploaded related videos.

[0089] As an optional example, the above-mentioned construction module includes: a construction unit, used to determine the association relationship between the first video and the associated video, wherein the association relationship includes a parallel relationship and a parent-child relationship; when the association relationship is a parallel relationship, a parallel label relationship tree is generated; when the association relationship is a parent-child relationship, a parent-child label relationship tree is generated.

[0090] In this embodiment, when the first video contains an associated video, it has been determined that the first video and the associated video need to construct a tag relationship tree. At this time, it is necessary to determine the association relationship between the first video and the associated video, whether it is a parent-child relationship or a parallel relationship. Among them, the parent-child relationship or the parallel relationship does not refer to which of the two videos contains which one, or the two are parallel, but the display order of the first video and the associated video in the collection is determined by the parent-child relationship and the parallel relationship. For the parent-child relationship, for example, a parent video is followed by two child videos, and the playback order of the two child videos is after the parent video. If the two child videos are in a parallel relationship, the playback order can be reversed.

[0091] As an optional example, the above-mentioned construction unit includes: a determination subunit, used to determine the associated information associated with the first video through the associated tags in the video information of the first video; or, obtain the title information in the video information of the first video, and determine, from multiple videos according to the title information, a video with a high degree of similarity between the title and the title information and with a serial number, and use the determined video as the associated video of the first video.

[0092] In this embodiment, when determining the association relationship between the first video and the associated videos, if the user determines the video tag, the associated videos related to the first video have been determined through the video tag of the first video, and the association relationship between the first video and the associated videos is marked as a parent-child relationship or a parallel relationship. If the system identifies the association relationship, the system can identify the first video and multiple videos, recognize that the title of the first video is highly similar to the titles of the multiple videos, and the first video and the multiple videos are numbered. At this time, it can be determined that the first video is associated with the multiple videos, and the association relationship is determined based on the serial number. The parent video is in the first serial number, the child video is in the second serial number, and the parallel relationship is the same for the serial numbers.

[0093] As an optional example, the above-mentioned determination module includes: a determination unit, used to determine the user intention based on the keywords in the video search request; determine a video that is the same or similar to the user intention from the video information of the video; and use the determined video as the target video.

[0094] In this embodiment, when the target user initiates a video search request, keywords can be extracted from the video search request, and then the user intent matching the keyword can be found based on the keyword. This process can be determined based on the neural network model. By inputting keywords, the neural network model gives the user intent, and analyzes the user intent to obtain the video that the user may want to search for. Among the results output by the model, the video with a higher possibility is used as the target video.

[0095] As an optional example, the second acquisition module includes: a traversal unit, which is used to, when the video tag of the target video is located in the tag relationship tree, start from the video tag of the target video, traverse the parent node upward, and the parent node of each parent node, until the root node is traversed; start from the root node, traverse each child node, and the child node of each child node, until all child nodes that do not contain any child nodes are traversed; and use the video corresponding to the tag of each traversed node as one of all videos.

[0096] In this embodiment, after acquiring the target video, the video tag of the target video is determined, and then the tag is searched from all tag relationship trees. After the relationship tree of the tag is found, traversal is performed to traverse all tags in the relationship tree.

[0097] The traversal method starts from the tag of the target video found, traverses upward to the root node, and then traverses downward from the root node to each node that does not contain child nodes. Thus, all nodes in each tag relationship tree are traversed. Each tag relationship tree traverses to obtain multiple video tags, and the multiple videos corresponding to the multiple video tags under the tag relationship tree constitute a video collection. Multiple tag relationship trees constitute multiple video collections.

[0098] In this embodiment, when the target user wants to search for videos and initiates a video search request, the server searches for multiple target videos based on the video search request, but does not recommend the target videos to the target user. Instead, the server recommends a video collection containing each target video among the multiple target videos to the target user.

[0099] As an optional example, the push module includes: a push unit, which is used to sort all videos according to the video information of each video in all videos; and push the video identifiers of all sorted videos to the target user.

[0100] In this embodiment, when each video collection is recommended to the user, the videos in each video collection must be sorted and then recommended to the user after sorting.

[0101] In this embodiment, when each video collection is recommended to the user, the videos in each video collection must be sorted and then recommended to the user after sorting.

[0102] There are many ways to sort, and the sorting order can be determined according to the parent-child relationship or parallel relationship in the tag relationship tree. One sorting method is to start from the root node, with the root node as the first video, and then traverse the child nodes of the root node. The videos of the child nodes are located after the root node. If there are multiple child nodes, the order of the multiple child nodes is parallel, and they can be sorted randomly or according to the order of video upload time. The child nodes continue to be traversed until the last node is traversed.

[0103] For example, Figure 4 As shown, Figure 4 This is a schematic diagram of different sorting results for the same tag relationship tree in this embodiment. For root node 1, there is only one, and then traverse the child nodes. If there are multiple child nodes, they are in a parallel relationship. The parallel relationship is randomly sorted. For example, the first and second cases are sorted as sort 2 and sort 5, and the third and fourth cases are sorted as sort 3 and sort 2. For non-parallel cases, such as sort 4 in the first case, there is only one child node, so it is directly sorted after sort 4, which is sort 5. After sorting, 6 videos are composed of Figure 4 There are four sorting situations in .

[0104] Another sorting method is to perform comprehensive sorting based on the relevance between the keywords provided by the user during the search and the recalled videos (from high to low), the upload time of the recalled videos (from recent to distant), and other sorting strategies (the purpose is to put the videos with the strongest relevance to the user's search intention in the first place, while also displaying other videos in the series to the user at the same time to meet the user's search needs and search experience).

[0105] This application first uses the video tags mounted on the video, for example, the tags can be video series identifiers (Identity document, id), to establish a parent-child series relationship tree (parent-child tag relationship tree) and a peer series relationship tree (parallel tag relationship tree); then, for the search request input by the target user, the series intention is determined and the corresponding series id is obtained; then, the parent-child series relationship tree and the peer series relationship tree are traversed and queried through the series id to obtain the corresponding main series id (root node id); then, all videos corresponding to the series id under the parent-child series relationship tree or the peer series relationship tree are recalled according to the main series id; finally, all recalled videos are sorted on the basis of full recall. After sorting, they are pushed to users.

[0106] Constructing a series tree: Using the series identification code (series ID) on the video document, establish a parent-child series relationship tree and a peer series relationship tree.

[0107] Identify query intent: Analyze and determine the query request entered by the user, find out its series intent, and obtain its corresponding series ID;

[0108] Traverse the series tree: Use the series ID to traverse the series relationship tree and find the main series ID;

[0109] Video recall: Based on the main series ID, all related series videos are recalled together;

[0110] Result sorting: Sort the recalled videos so that users can find the videos they need more conveniently and quickly.

[0111] When uploading each video, we check whether the server has stored related videos of the video. If there are related videos, we build a tag relationship tree. If there are no related videos, the uploaded video can form a series. After building the tag relationship tree, if the user searches for any video while watching TV, the entire series will be recommended to the user.

[0112] For example, when a user uploads the second episode of the third season of a children's program, it checks whether the content of other seasons of the children's program and other episodes of the third season are stored. If they are stored, a label relationship tree is established. Programs of different seasons are in a parallel relationship and are located under the same parent node. Different episodes of the same season are in a parallel relationship and are located under the same season point. If the user's search keyword hits any episode of any season of the children's program, the label relationship tree is traversed to find the content of each episode of each season of the children's program stored on the server, and then sent to the user (sending identification) after sorting for the user to choose.

[0113] This embodiment organizes and indexes the video content under the series through a series tree structure. The series tree structure can effectively reflect the series relationship between videos, including parent-child series relationship and peer series relationship. This structure can make the recall of all video content very efficient. Even if the user's query content is slightly biased, this application takes all the video content under the relevant series into consideration through the series tree traversal, thereby greatly enhancing the accuracy and comprehensiveness of the search. It improves the problems of low accuracy of search results and incomplete search scope that are common in current network searches.

[0114] like Figure 6 As shown, an embodiment of the present application provides an electronic device, including a processor 111, a communication interface 112, a memory 113 and a communication bus 114, wherein the processor 111, the communication interface 112, and the memory 113 communicate with each other through the communication bus 114.

[0115] Memory 113, used for storing computer programs;

[0116] In one embodiment of the present application, the processor 111 is used to implement the video recall method provided by any one of the aforementioned method embodiments when executing the program stored in the memory 113.

[0117] An embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, a video recall method provided in any of the aforementioned method embodiments is implemented.

[0118] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or may be distributed over multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0119] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment.

[0120] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "" as used herein may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive and therefore specify the presence of stated features, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0121] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features applied herein.

Claims

1. A video recall method, characterized in that: include: Upon receiving a video upload request, obtaining a first video requested to be uploaded by the video upload request and a video tag of the first video; When the first video includes associated associated videos, constructing a tag relationship tree according to the video tag of the first video and the video tags of the associated videos; When receiving a video search request from a target user, determining a searched target video according to the video search request; When the video tag of the target video is in the tag relationship tree, obtaining all videos corresponding to all tags in the tag relationship tree; The video identifiers of all the videos are pushed to the target user.

2. The method according to claim 1, characterized in that Obtaining the video tag of the first video includes: Acquire a video category of the first video from the video information of the first video; A video tag is generated for the first video according to the video classification, wherein the video tag is used to uniquely identify the first video, and the video tag indicates the video classification of the first video.

3. The method according to claim 1, characterized in that When the first video includes associated associated videos, constructing a tag relationship tree according to the video tag of the first video and the video tags of the associated videos includes: Determine an association relationship between the first video and the associated video, wherein the association relationship includes a parallel relationship and a parent-child relationship; When the association relationship is a parallel relationship, generating a parallel label relationship tree; When the association relationship is a parent-child relationship, a parent-child label relationship tree is generated.

4. The method according to claim 3, characterized in that Determining the association relationship between the first video and the associated video includes: Determine the association relationship with the associated video through the association tag in the video information of the first video; or, Obtain title information in the video information of the first video, and determine a video with a title that has a high similarity with the title information and has a serial number from multiple videos based on the title information, use the determined video as the first video and determine the association relationship.

5. The method according to claim 1, characterized in that When receiving a video search request from a target user, determining the searched target video according to the video search request includes: Determining user intent based on keywords in the video search request; Determine, from the video information of the video, a video that is the same as or similar to the user's intention; The determined video is used as the target video.

6. The method according to claim 1, characterized in that In a case where the video tag of the target video is located in the tag relationship tree, obtaining all videos corresponding to all tags in the tag relationship tree includes: When the video tag of the target video is located in the tag relationship tree, starting from the video tag of the target video, traversing upward through the parent nodes and the parent nodes of each parent node until the root node is traversed; Starting from the root node, traverse each child node, and each child node's child node, until all child nodes that do not contain any child nodes are traversed; The video corresponding to the label of each traversed node is taken as one of all the videos.

7. The method according to claim 1, characterized in that Pushing video IDs of all videos to target users includes: Sort all videos according to the video information of each video; The video identifiers of all the sorted videos are pushed to the target user.

8. A video recall device, characterized in that: include: An acquisition module, configured to acquire, upon receiving a video upload request, a first video requested to be uploaded and a video tag of the first video; A construction module, configured to construct a tag relationship tree according to the video tag of the first video and the video tags of the associated videos when the first video includes associated associated videos; A determination module, configured to determine the searched target video according to the video search request when receiving a video search request from a target user; An acquisition module, configured to acquire all videos corresponding to all tags in the tag relationship tree when the video tag of the target video is located in the tag relationship tree; The push module is used to push the video identifiers of all the videos to the target user.

9. An electronic device, characterized in that: include: at least one communication interface; at least one bus connected to at least one communication interface; at least one processor coupled to at least one bus; At least one memory connected to at least one bus, wherein a computer program is stored in the memory, and when a processor executes the computer program, the video recall method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are used to execute the video recall method of any one of claims 1 to 7 of the present application.

Images