A content recommendation method and system for short video users

Abstract

The present application relates to the field of video processing technology, and in particular to a content recommendation method and system for short video users, which collects a list of highlight short videos; outputs a list of popular short videos and a list of unpopular short videos through the highlight metadata of each highlight short video in the highlight short video list and a designated short video database; for the unpopular short videos in the unpopular short video list, compares the highlight segments of the unpopular short videos with the highlight segments of the popular short videos in the hot short video list, and selects the hot short video with the highest similarity of the highlight segments as the recommended short video for the unpopular short video. The present invention accurately segments and locates short videos of the nature of highlights, and adaptively matches and associates the current popular and unpopular short videos in combination with the historical viewing needs of the user, and makes accurate recommendations to the user.

Description

Technical Field

[0001] The present application relates to the field of video processing technology, and in particular to a content recommendation method and system for short video users. Background Art

[0002] With the development of computer and network technology, the functions of electronic devices are becoming more and more diversified. It is becoming more and more popular for users to splice interesting video segments into new videos by video editing.

[0003] In the prior art, with the continuous development of smart terminal devices, users have an increasing demand for the production of highlight videos. In particular, highlight videos that display multiple videos in a collection and connection, accompanied by commentary and explanation, are becoming more and more popular.

[0004] However, it is still difficult to evaluate the quality of numerous highlight videos and find the highlight video that best suits user needs and objective quality. Summary of the invention

[0005] To achieve the above objectives, this application provides the following technical solutions:

[0006] According to a first aspect of the present invention, the present invention claims protection for a content recommendation method for short video users, comprising:

[0007] Collection of short video lists;

[0008] Outputting a hot short video list and a cold short video list through the highlight metadata of each highlight short video in the highlight short video list and a designated short video database;

[0009] For the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list, and the popular short video with the highest similarity in the highlight segments is selected as the recommended short video for the unpopular short video.

[0010] Furthermore, for the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list, and the popular short video with the highest similarity in the highlight segments is selected as the recommended short video for the unpopular short video specifically includes:

[0011] For the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list;

[0012] Selecting a popular short video having the largest number of highlight segments compared with the highlight segments split from the unpopular short video as a recommended short video for the unpopular short video;

[0013] Or, for the unpopular short videos in the unpopular short video list, comparing the highlight segments split from the unpopular short videos with the highlight segments split from the popular short videos in the popular short video list, and selecting the popular short video with the highest similarity in the highlight segments as the recommended short video for the unpopular short video specifically includes:

[0014] For the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list;

[0015] According to the order of the number of highlight clips compared with the unpopular short video from high to low, a preset number of popular short videos are selected as the hot short videos to be split from the unpopular short video;

[0016] Outputting the similarity between the unpopular short video and each of the hot short videos to be split according to the number of the compared highlight segments and the number of the highlight segments split from the hot short videos to be split;

[0017] The popular short video to be split with the highest similarity is selected as the recommended short video for the unpopular short video.

[0018] Furthermore, the method further comprises:

[0019] Analyze whether the highlights of unpopular short videos are of the same type as the highlights of popular short videos split from the unpopular short videos;

[0020] If the types are consistent, then analyze whether the highlight elements and their order of the highlight contents of the two are consistent. If one highlight element is different, mark the unpopular short video as the preferred highlight element short video of the split popular short video; if the highlight elements are completely consistent but the order of the highlight elements is not completely consistent, then mark the unpopular short video as a short video with similar highlight elements of the split popular short video; if multiple highlight elements are inconsistent, check whether the highlight contents of the two exist in the same highlight metadata. If so, mark the unpopular short video as an extended short video of the recommended short video; if not, analyze whether the similarity of the highlight clips of the two is greater than the specified similarity threshold. If so, mark the unpopular short video as a variant short video of the recommended short video;

[0021] If the types are inconsistent, analyze whether the highlight elements of the two highlight contents have one more highlight element or one less highlight element and the other highlight elements are completely consistent. If so, mark the unpopular short video as a short video with more highlight elements or fewer highlight elements of the split popular short video; if not, check whether the highlight contents of the two exist in the same highlight metadata. If so, mark the unpopular short video as an extended short video of the split popular short video; if not, analyze whether the similarity of the highlight clips of the two is greater than the specified similarity threshold. If so, mark the unpopular short video as a variant short video of the split popular short video.

[0022] Furthermore, for the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list as follows:

[0023] For unpopular short videos in the unpopular short video list whose number of split highlight segments is greater than or equal to a first threshold, the split highlight segments are compared with the highlight segments of popular short videos in the popular short video list whose number of split highlight segments is greater than or equal to a second threshold;

[0024] The similarity between the unpopular short video and each of the hot short videos to be split is outputted by the number of the compared highlight segments and the number of the highlight segments split from the hot short videos to be split, specifically:

[0025] The ratio of the number of compared highlight clips to the number of highlight clips split from the hot short video to be split is calculated based on the number of compared highlight clips and the number of highlight clips split from the hot short video to be split, and the ratio is used as the similarity between the unpopular short video and the hot short video to be split.

[0026] Furthermore, the outputting of the hot short video list and the unpopular short video list by using the highlight metadata of each highlight short video in the highlight short video list and the designated short video database specifically includes:

[0027] Classifying the highlight metadata of each highlight short video in the highlight short video list, and outputting the highlight attribute information and the short video highlight content of each highlight metadata;

[0028] Compare the short video collection content of each collection metadata with the collection content of the popular short videos or the image attributes of the collection content in the corresponding collection attributes in the specified short video database, and output the popular short videos and the unpopular short videos;

[0029] Splicing the hot short videos and their split highlight clips that are consistent among all the hot short videos, and forming a hot short video list with all the hot short videos and their split highlight clips that are output after the splicing operation;

[0030] The unpopular short videos and their split highlight segments that are consistent among all the unpopular short videos are spliced ​​together, and all the unpopular short videos and their split highlight segments that are output after the splicing operation form an unpopular short video list.

[0031] According to a second aspect of the present invention, the present invention claims protection for a content recommendation system for short video users, comprising:

[0032] The collection module is used to collect the list of short videos;

[0033] A classification module, used to output a hot short video list and a cold short video list according to the highlight metadata of each highlight short video in the highlight short video list and a specified short video database;

[0034] The comparison and selection module is used to compare the highlight segments split from the unpopular short videos in the unpopular short video list with the highlight segments split from the popular short videos in the popular short video list, and select the popular short video with the highest similarity in the highlight segments as the recommended short video for the unpopular short video.

[0035] Furthermore, the comparison and selection module specifically includes:

[0036] A comparison submodule, for comparing, for the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos with the highlight segments split from the popular short videos in the popular short video list;

[0037] A first selection submodule is used to select a popular short video having the largest number of highlight segments compared with the highlight segments split from the unpopular short video as a recommended short video for the unpopular short video;

[0038] The comparison and selection module specifically includes:

[0039] A comparison submodule, for comparing, for the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos with the highlight segments split from the popular short videos in the popular short video list;

[0040] The second selection submodule is used to select a preset number of popular short videos as the hot short videos to be split from the unpopular short video in descending order of the number of highlight clips compared with the unpopular short video;

[0041] A calculation submodule, used for outputting the similarity between the unpopular short video and each of the hot short videos to be split according to the number of the compared highlight segments and the number of the highlight segments split from the hot short videos to be split;

[0042] The third selection submodule is used to select the hot short video to be split with the highest similarity as the recommended short video for the unpopular short video.

[0043] Furthermore, the system further comprises:

[0044] The first analysis module is used to analyze whether the types of the highlights of the unpopular short video and the highlights of the popular short video split from the unpopular short video are consistent;

[0045] The second parsing module is used to, when the parsing result of the first parsing module is of the same type, parse whether the highlight elements and their order of the highlight contents of the two are consistent; if one highlight element is different, mark the unpopular short video as a preferred highlight element short video of the split popular short video; if the highlight elements are completely consistent but the order of the highlight elements is not completely consistent, mark the unpopular short video as a short video with similar highlight elements of the split popular short video; if multiple highlight elements are inconsistent, check whether the highlight contents of the two exist in the same highlight metadata; if so, mark the unpopular short video as an extended short video of the recommended short video; if not, parse whether the similarity of the highlight clips of the two is greater than a specified similarity threshold; if so, mark the unpopular short video as a variant short video of the recommended short video;

[0046] The third parsing module is used to, when the parsing result of the first parsing module is inconsistent in type, parse whether the highlight elements of the highlight contents of the two have one more highlight element or one less highlight element and the other highlight elements are completely consistent; if so, mark the unpopular short video as a short video with more highlight elements or fewer highlight elements of the split popular short video; if not, check whether the highlight contents of the two exist in the same highlight metadata; if so, mark the unpopular short video as an extended short video of the split popular short video; if not, parse whether the similarity of the highlight clips of the two is greater than a specified similarity threshold; if so, mark the unpopular short video as a variant short video of the split popular short video.

[0047] Furthermore, the comparison submodule is specifically used to compare the highlight segments of the unpopular short videos in the unpopular short video list whose number of highlight segments is greater than or equal to a first threshold with the highlight segments of the popular short videos in the popular short video list whose number of highlight segments is greater than or equal to a second threshold;

[0048] The calculation submodule is specifically used to calculate the ratio of the number of compared highlight clips to the number of highlight clips split from the hot short video to be split, and use the ratio as the similarity between the unpopular short video and the hot short video to be split.

[0049] Furthermore, the classification module specifically includes:

[0050] A grading submodule, for grading the highlight metadata of each highlight short video in the highlight short video list, and outputting the highlight attribute information and the short video highlight content of each highlight metadata;

[0051] The short video comparison submodule is used to compare the short video collection content of each collection metadata with the collection content of the hot short videos or the image attributes of the collection content in the corresponding collection attributes in the specified short video database, and output the hot short videos and the unpopular short videos;

[0052] The first splicing submodule is used to splice the hot short videos and their split highlight clips that are consistent among all the hot short videos, and form a hot short video list with all the hot short videos and their split highlight clips output after the splicing operation;

[0053] The second splicing submodule is used to splice the consistent unpopular short videos and their split highlight clips among all the unpopular short videos, and form an unpopular short video list with all the unpopular short videos and their split highlight clips output after the splicing operation.

[0054] The present application relates to the field of video processing technology, and in particular to a content recommendation method and system for short video users, which collects a list of highlight short videos; outputs a list of popular short videos and a list of unpopular short videos through the highlight metadata of each highlight short video in the highlight short video list and a designated short video database; for the unpopular short videos in the unpopular short video list, compares the highlight segments of the unpopular short videos with the highlight segments of the popular short videos in the hot short video list, and selects the hot short video with the highest similarity of the highlight segments as the recommended short video for the unpopular short video. The present invention accurately segments and locates short videos of the nature of highlights, and adaptively matches and associates the current popular and unpopular short videos in combination with the historical viewing needs of the user, and makes accurate recommendations to the user. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A workflow diagram of a content recommendation method for short video users as claimed in an embodiment of the present application;

[0056] Figure 2 A second workflow diagram of a content recommendation method for short video users claimed in an embodiment of the present application;

[0057] Figure 3 A schematic diagram of the structure of a content recommendation system for short video users claimed in an embodiment of the present application;

[0058] Figure 4 A schematic diagram of the structure of a comparison and selection module of a content recommendation system for short video users claimed in an embodiment of the present application;

[0059] Figure 5 A second structural diagram of a content recommendation system for short video users as claimed in an embodiment of the present application. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only 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 creative work are within the scope of protection of this application.

[0061] The terms "first", "second" and "third" in this application are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined as "first", "second" and "third" can explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. In the embodiments of this application, all directional indications (such as up, down, left, right, front, back...) are only used to explain the relative position relationship, movement, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication also changes accordingly. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or modules is not limited to the listed steps or modules, but optionally also includes steps or modules that are not listed, or optionally also includes other steps or modules inherent to these processes, methods, products or devices.

[0062] Reference to "embodiment" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiment may be included in at least one embodiment of the present application. The appearance of this phrase in various locations in the specification does not necessarily refer to a consistent embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0063] See also Figure 1 , which is a flow chart of Example 1 of a content recommendation method for short video users according to an embodiment of the present invention, Figure 1 It can be seen that the method specifically includes:

[0064] S101: Collect the short video list;

[0065] In the specific implementation, if the highlight short video database stores the highlight short videos independently, a certain number of highlight short videos can be collected from the highlight short video database and spliced ​​into a highlight short video list; if the highlight short video database stores the highlight short videos in a list format, the highlight short video list can also be directly collected from the highlight short video library.

[0066] S102: Outputting a popular short video list and a less popular short video list through the highlight metadata of each highlight short video in the highlight short video list and a designated short video database;

[0067] Specifically, this step can be implemented in the following ways:

[0068] Classifying the highlight metadata of each highlight short video in the highlight short video list, and outputting the highlight attribute information and the short video highlight content of each highlight metadata;

[0069] Compare the short video collection content of each collection metadata with the collection content of the popular short videos or the image attributes of the collection content in the corresponding collection attributes in the specified short video database, and output the popular short videos and the unpopular short videos;

[0070] Splicing the hot short videos and their split highlight clips that are consistent among all the hot short videos, and forming a hot short video list with all the hot short videos and their split highlight clips that are output after the splicing operation;

[0071] The unpopular short videos and their split highlight segments that are consistent among all the unpopular short videos are spliced ​​together, and all the unpopular short videos and their split highlight segments that are output after the splicing operation form an unpopular short video list.

[0072] Among them, the "designated short video database" can be a short video database established through short video information provided by the short video platform information system, and the short videos stored in the designated short video database are all popular short videos. The highlight metadata is graded, and the highlight attribute information and short video information of the highlight metadata are output. The short video highlight content of each highlight metadata is compared with the highlight content of the popular short videos in the corresponding highlight attribute in the designated short video database or the image attributes of the highlight content. If the comparison is consistent, it is determined that the short video with the highlight metadata is a popular short video. If the comparison is inconsistent, it is determined that the short video with the highlight metadata is a unpopular short video. Compare and output the popular short videos and the unpopular short videos, and then output the popular short video list and the unpopular short video list respectively by splicing the short videos and their split highlight segments;

[0073] The highlight metadata represents metadata of a highlight video, such as the number of highlight videos of the video, the highlight type, etc.;

[0074] The classification of the metadata of the collection shall at least include the classification of the scale and the age of the audience group according to the corresponding value and content of the specific metadata. For example, if the number of the collection videos is greater than 2,000 and the collection type is a romantic short video, the level shall be set as a high-view adult-level video;

[0075] The highlight attribute information and short video information of the highlight metadata are the corresponding contents of the metadata, such as the number of highlight videos and the corresponding type label of the highlight type.

[0076] 103: For the unpopular short videos in the unpopular short video list, compare the highlight segments split from the unpopular short videos with the highlight segments split from the popular short videos in the popular short video list, and select the popular short video with the highest similarity in the highlight segments as the recommended short video for the unpopular short video.

[0077] In specific implementation, this step can be implemented in the following ways, including:

[0078] For the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list;

[0079] A popular short video having the largest number of highlight segments compared with the highlight segments split from the unpopular short video is selected as a recommended short video for the unpopular short video.

[0080] In specific implementation, this step can also be implemented in the following ways:

[0081] For the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list;

[0082] According to the order of the number of highlight clips compared with the unpopular short video from high to low, a preset number of popular short videos are selected as the hot short videos to be split from the unpopular short video;

[0083] Outputting the similarity between the unpopular short video and each of the hot short videos to be split according to the number of the compared highlight segments and the number of the highlight segments split from the hot short videos to be split;

[0084] The popular short video to be split with the highest similarity is selected as the recommended short video for the unpopular short video.

[0085] Specifically, the above-mentioned "output the similarity between the unpopular short video and each hot short video to be split according to the number of compared highlight clips and the number of highlight clips split from the hot short video to be split" is specifically implemented by calculating the ratio of the number of compared highlight clips to the number of highlight clips split from the hot short video to be split according to the number of compared highlight clips and the number of highlight clips split from the hot short video to be split, and using the ratio as the similarity between the unpopular short video and the hot short video to be split.

[0086] In specific implementation, in order to improve the accuracy of the analysis of unpopular short videos in the highlight metadata, the present invention can also restrict the unpopular short videos and popular short videos involved in the comparison. Specifically, the above "for unpopular short videos in the unpopular short video list, compare the highlight segments split from the unpopular short videos with the highlight segments split from the popular short videos in the popular short video list" is specifically:

[0087] For unpopular short videos in the unpopular short video list whose number of split highlight segments is greater than or equal to a first threshold, the split highlight segments are compared with the highlight segments of popular short videos in the popular short video list whose number of split highlight segments is greater than or equal to a second threshold. The first threshold and the second threshold are divided based on the specific duration of the short video.

[0088] It can be seen from the above embodiments that the present invention collects a list of highlights short videos, and outputs a list of popular short videos and a list of unpopular short videos through the highlights metadata of each highlight short video in the list of highlights short videos and a designated short video database; for the unpopular short videos in the list of unpopular short videos, the highlights segments of the unpopular short videos are compared with the highlights segments of the popular short videos in the list of popular short videos, and the popular short videos with the highest similarity of the highlights segments are selected as the recommended short videos for the unpopular short videos. Since the highlights segments of the short videos are highlights segments that have a certain split with the short videos in the short video platform position, the present invention analyzes the splitting relationship between the unpopular short videos and the hot short videos by analyzing the comparison degree of the highlights segments of the unpopular short videos and the hot short videos. Its essence is to analyze the relationship between the unpopular short videos and the hot short videos from the perspective of the short video platform position, which can improve the accuracy of short video recognition in the highlights, lay a good foundation for other businesses based on short video information, and provide convenience for users.

[0089] Embodiment 2

[0090] In practical applications, in addition to correctly identifying short videos in the collection metadata, it is also necessary to further analyze the correspondence between unpopular short videos and recommended short videos. Based on this, the present invention also provides a preferred solution. This preferred solution is based on the above-mentioned embodiment one, and adds an operating method for parsing the correspondence between unpopular short videos and split popular short videos by utilizing the collection content of unpopular short videos and the collection content type.

[0091] See also Figure 2 , which is a flow chart of Example 2 of a content recommendation method for short video users according to an embodiment of the present invention, Figure 2 It can be seen that the method specifically includes:

[0092] S201: Collect a list of highlights short videos;

[0093] S202: Outputting a popular short video list and a less popular short video list through the highlight metadata of each highlight short video in the highlight short video list and a designated short video database;

[0094] S203: For the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list, and the popular short video with the highest similarity in the highlight segments is selected as the recommended short video for the unpopular short video.

[0095] The above S201 to S203 are consistent with S101 to S103 in the first embodiment and will not be described in detail here.

[0096] S204: analyzing whether the types of the highlights of the unpopular short video and the highlights of the popular short video split from the unpopular short video are consistent; if they are consistent, proceeding to S205; if not, proceeding to S206;

[0097] S205: Analyze whether the highlight elements and their order of the highlight contents of the two are consistent. If one highlight element is different, mark the unpopular short video as a preferred highlight element short video of the split popular short video; if the highlight elements are completely consistent but the order of the highlight elements is not completely consistent, mark the unpopular short video as a short video with similar highlight elements of the split popular short video; if multiple highlight elements are inconsistent, check whether the highlight contents of the two exist in the same highlight metadata. If so, mark the unpopular short video as an extended short video of the recommended short video; if not, analyze whether the similarity of the highlight clips of the two is greater than a specified similarity threshold. If so, mark the unpopular short video as a variant short video of the recommended short video;

[0098] S206: Analyze whether the highlight elements of the highlight contents of the two have one more highlight element or one less highlight element and the other highlight elements are completely consistent. If so, mark the unpopular short video as a short video with more highlight elements or fewer highlight elements of the split popular short video; if not, check whether the highlight contents of the two exist in the same highlight metadata. If so, mark the unpopular short video as an extended short video of the split popular short video; if not, analyze whether the similarity of the highlight clips of the two is greater than a specified similarity threshold. If so, mark the unpopular short video as a variant short video of the split popular short video.

[0099] In a specific implementation, the value range of the similarity threshold may be greater than 60%; preferably, the similarity threshold is 80%.

[0100] For example, the unpopular short video is "Collection of Qing Dynasty Customs" and the popular short video is "Collection of Zhen Huan Zhuan". If S201-S203 analyzes that the unpopular short video is separated from the popular short video, and then S204-S205 analyzes that the unpopular short video is the preferred collection element short video of the popular short video;

[0101] The unpopular short video is a "collection of love poems", and the popular short video is a "collection of love songs". If the unpopular short video is separated from the popular short video through S201-S203 analysis, and then the unpopular short video is analyzed through S204-S206 to be a short video with less collection elements of the popular short video;

[0102] The unpopular short video is "Collection of Parent-child Relationship Issues", and the popular short video is "Collection of Family with Children". If the unpopular short video is separated from the popular short video through S201-S203 analysis, and then the unpopular short video is analyzed through S204-S205 to be a short video with similar elements to the collection of popular short videos;

[0103] Furthermore, the above method may also include: displaying prompt information including the correspondence between unpopular short videos and recommended short videos, and modifying the highlight metadata according to the recommended short video selected by the user.

[0104] It can be seen from the above embodiments that after analyzing the relationship between unpopular short videos and popular short videos, the present invention also analyzes the relationship between the collection content types, collection elements of the collection content and the order of the unpopular short videos and the recommended short videos, and marks the corresponding relationships between the unpopular short videos and the recommended short videos, such as preferred collection elements, fewer collection elements, more collection elements, similarity of collection elements, and variation.

[0105] Embodiment 3

[0106] Corresponding to the above-mentioned content recommendation method for short video users, an embodiment of the present invention also provides a content recommendation system for short video users.

[0107] See also Figure 3 , Figure 3 1 is a structural diagram of Example 1 of a content recommendation system for short video users according to an embodiment of the present invention; Figure 3 It can be seen that the system specifically includes:

[0108] A collection module 301 is used to collect a list of short videos;

[0109] The classification module 302 is used to output a hot short video list and a cold short video list according to the collection metadata of each collection short video in the collection short video list and a designated short video database;

[0110] The comparison and selection module 303 is used to compare the highlight segments split from the unpopular short videos in the unpopular short video list with the highlight segments split from the popular short videos in the popular short video list, and select the popular short video with the highest similarity in the highlight segments as the recommended short video for the unpopular short video.

[0111] like Figure 4 As shown, comparing the selected modules may include:

[0112] A comparison submodule 401 is used to compare, for the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos with the highlight segments split from the popular short videos in the popular short video list;

[0113] A first selection submodule 402 is used to select a popular short video having the largest number of highlight segments compared with the highlight segments split from the unpopular short video as a recommended short video for the unpopular short video;

[0114] The second selection submodule 403 is used to select a preset number of popular short videos as the hot short videos to be split from the unpopular short video in descending order of the number of highlight clips compared with the unpopular short video;

[0115] The calculation submodule 404 is used to output the similarity between the unpopular short video and each of the hot short videos to be split according to the number of the compared highlight segments and the number of the highlight segments split from the hot short videos to be split;

[0116] The third selection submodule 405 is used to select the hot short video to be split with the highest similarity as the recommended short video for the unpopular short video.

[0117] Preferably, the above-mentioned comparison submodule is specifically used to compare the highlight segments split from unpopular short videos in the unpopular short video list, whose number of highlight segments split is greater than or equal to a first threshold, with the highlight segments of popular short videos in the popular short video list, whose number of highlight segments split is greater than or equal to a second threshold.

[0118] Preferably, the above-mentioned calculation submodule is specifically used to calculate the ratio of the number of compared highlight clips to the number of highlight clips split from the popular short video to be split, through the number of compared highlight clips and the number of highlight clips split from the popular short video to be split, and use the ratio as the similarity between the unpopular short video and the popular short video to be split.

[0119] Preferably, the above classification module specifically includes:

[0120] A grading submodule, for grading the highlight metadata of each highlight short video in the highlight short video list, and outputting the highlight attribute information and the short video highlight content of each highlight metadata;

[0121] The short video comparison submodule is used to compare the short video collection content of each collection metadata with the collection content of the hot short videos or the image attributes of the collection content in the corresponding collection attributes in the specified short video database, and output the hot short videos and the unpopular short videos;

[0122] The first splicing submodule is used to splice the hot short videos and their split highlight clips that are consistent among all the hot short videos, and form a hot short video list with all the hot short videos and their split highlight clips output after the splicing operation;

[0123] The second splicing submodule is used to splice the consistent unpopular short videos and their split highlight clips among all the unpopular short videos, and form an unpopular short video list with all the unpopular short videos and their split highlight clips output after the splicing operation.

[0124] like Figure 5 As shown, Figure 5 This is a structural diagram of Example 2 of a content recommendation system for short video users according to an embodiment of the present invention. The system may further include, based on the system of Example 1 above:

[0125] The first analysis module 304 is used to analyze whether the types of the highlights of the unpopular short video and the highlights of the popular short videos split from the unpopular short video are consistent;

[0126] The second parsing module 305 is used to, when the parsing result of the first parsing module is that the type is consistent, parse whether the highlight elements and their order of the highlight contents of the two are consistent. If one highlight element is different, mark the unpopular short video as a preferred highlight element short video of the split popular short video; if the highlight elements are completely consistent but the order of the highlight elements is not completely consistent, mark the unpopular short video as a short video with similar highlight elements of the split popular short video; if multiple highlight elements are inconsistent, check whether the highlight contents of the two exist in the same highlight metadata. If so, mark the unpopular short video as an extended short video of the recommended short video; if not, parse whether the similarity of the highlight clips of the two is greater than a specified similarity threshold. If so, mark the unpopular short video as a variant short video of the recommended short video;

[0127] The third parsing module 306 is used to, when the parsing result of the first parsing module is inconsistent in type, analyze whether the highlight elements of the highlight contents of the two have one more highlight element or one less highlight element and the other highlight elements are completely consistent. If so, mark the unpopular short video as a short video with more highlight elements or fewer highlight elements of the split popular short video; if not, check whether the highlight contents of the two exist in the same highlight metadata. If so, mark the unpopular short video as an extended short video of the split popular short video; if not, analyze whether the similarity of the highlight clips of the two is greater than a specified similarity threshold. If so, mark the unpopular short video as a variant short video of the split popular short video.

[0128] It can be seen from the above embodiments of the present invention that the system provided by the present invention, by collecting a list of highlights short videos, outputs a list of popular short videos and a list of unpopular short videos through the highlights metadata of each highlight short video in the list of highlights short videos and a designated short video database; for the unpopular short videos in the list of unpopular short videos, the highlights segments of the unpopular short videos are compared with the highlights segments of the popular short videos in the list of popular short videos, and the popular short videos with the highest similarity of the highlights segments are selected as the recommended short videos of the unpopular short videos. Since the highlights segments of the short videos are highlights segments that have a certain split with the short videos in the short video platform position, the present invention analyzes the splitting relationship between the unpopular short videos and the hot short videos by analyzing the comparison degree of the highlights segments of the unpopular short videos and the hot short videos. Its essence is to analyze the relationship between the unpopular short videos and the hot short videos from the perspective of the short video platform position, which can improve the accuracy of short video recognition in the highlights, lay a good foundation for other businesses based on short video information, and provide convenience for users.

[0129] In the several embodiments provided in the present application, it should be understood that the disclosed systems, systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the division of modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of systems or modules, which can be electrical, mechanical or other forms.

[0130] In addition, each functional module in each embodiment of the present application can be integrated into one operating module, or each module can exist physically separately, or two or more modules can be integrated into one module. The above integrated modules can be implemented in the form of hardware or in the form of software functional modules. The above is only an implementation method of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the specification and drawings of this application, or directly or indirectly used in other related technical fields, are also included in the patent protection scope of the present application.

[0131] The specific implementation methods of the invention are described in detail above, but they are only examples, and the present application is not limited to the specific implementation methods described above. For those skilled in the art, any equivalent modification or substitution of the invention is also within the scope of the present application, and therefore, the equalization, modification, and improvement made without departing from the spirit and principle of the present application should be included in the scope of the present application.

Claims

1. A content recommendation method for short video users, characterized in that: include: Collection of short video lists; Outputting a hot short video list and a cold short video list through the highlight metadata of each highlight short video in the highlight short video list and a designated short video database; For the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list, and the popular short video with the highest similarity in the highlight segments is selected as the recommended short video for the unpopular short video; The method further comprises: Analyze whether the highlights of unpopular short videos are of the same type as the highlights of popular short videos split from the unpopular short videos; If the types are consistent, then analyze whether the highlight elements and their order of the highlight contents of the two are consistent. If one highlight element is different, mark the unpopular short video as the preferred highlight element short video of the split popular short video; if the highlight elements are completely consistent but the order of the highlight elements is not completely consistent, then mark the unpopular short video as a short video with similar highlight elements of the split popular short video; if multiple highlight elements are inconsistent, check whether the highlight contents of the two exist in the same highlight metadata. If so, mark the unpopular short video as an extended short video of the recommended short video; if not, analyze whether the similarity of the highlight clips of the two is greater than the specified similarity threshold. If so, mark the unpopular short video as a variant short video of the recommended short video; If the types are inconsistent, analyze whether the highlight elements of the two highlight contents have one more highlight element or one less highlight element and the other highlight elements are completely consistent. If so, mark the unpopular short video as a short video with more highlight elements or fewer highlight elements of the split popular short video; if not, check whether the highlight contents of the two exist in the same highlight metadata. If so, mark the unpopular short video as an extended short video of the split popular short video; if not, analyze whether the similarity of the highlight clips of the two is greater than the specified similarity threshold. If so, mark the unpopular short video as a variant short video of the split popular short video.

2. The method according to claim 1, characterized in that The step of comparing the highlight segments of the unpopular short videos in the unpopular short video list with the highlight segments of the popular short videos in the popular short video list, and selecting the popular short video with the highest similarity of the highlight segments as the recommended short video for the unpopular short video specifically includes: For the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list; Selecting a popular short video having the largest number of highlight segments compared with the highlight segments split from the unpopular short video as a recommended short video for the unpopular short video; Or, for the unpopular short videos in the unpopular short video list, comparing the highlight segments split from the unpopular short videos with the highlight segments split from the popular short videos in the popular short video list, and selecting the popular short video with the highest similarity in the highlight segments as the recommended short video for the unpopular short video specifically includes: For the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list; According to the order of the number of highlight clips compared with the unpopular short video from high to low, a preset number of popular short videos are selected as the hot short videos to be split from the unpopular short video; Outputting the similarity between the unpopular short video and each of the hot short videos to be split according to the number of the compared highlight segments and the number of the highlight segments split from the hot short videos to be split; The popular short video to be split with the highest similarity is selected as the recommended short video for the unpopular short video.

3. The method according to claim 2, characterized in that For the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos are compared with the highlight segments split from the popular short videos in the popular short video list as follows: For unpopular short videos in the unpopular short video list whose number of split highlight segments is greater than or equal to a first threshold, the split highlight segments are compared with the highlight segments of popular short videos in the popular short video list whose number of split highlight segments is greater than or equal to a second threshold; The similarity between the unpopular short video and each of the hot short videos to be split is outputted based on the number of compared highlight segments and the number of highlight segments split from the hot short videos to be split, specifically: The ratio of the number of compared highlight clips to the number of highlight clips split from the hot short video to be split is calculated based on the number of compared highlight clips and the number of highlight clips split from the hot short video to be split, and the ratio is used as the similarity between the unpopular short video and the hot short video to be split.

4. The method according to claim 3, characterized in that Outputting a hot short video list and a cold short video list through the highlight metadata of each highlight short video in the highlight short video list and a designated short video database specifically includes: Classifying the highlight metadata of each highlight short video in the highlight short video list, and outputting the highlight attribute information and the short video highlight content of each highlight metadata; Compare the short video collection content of each collection metadata with the collection content of the popular short videos or the image attributes of the collection content in the corresponding collection attributes in the specified short video database, and output the popular short videos and the unpopular short videos; Splicing the hot short videos and their split highlight clips that are consistent among all the hot short videos, and forming a hot short video list with all the hot short videos and their split highlight clips that are output after the splicing operation; The unpopular short videos and their split highlight segments that are consistent among all the unpopular short videos are spliced ​​together, and all the unpopular short videos and their split highlight segments that are output after the splicing operation form an unpopular short video list.

5. A content recommendation system for short video users, characterized in that: include: The collection module is used to collect the list of short videos; A classification module, used to output a hot short video list and a cold short video list according to the highlight metadata of each highlight short video in the highlight short video list and a specified short video database; A comparison and selection module is used to compare the highlight segments of the unpopular short videos in the unpopular short video list with the highlight segments of the popular short videos in the popular short video list, and select the popular short video with the highest similarity of the highlight segments as the recommended short video of the unpopular short video; The system further comprises: The first analysis module is used to analyze whether the types of the highlights of the unpopular short video and the highlights of the popular short video split from the unpopular short video are consistent; The second parsing module is used to, when the parsing result of the first parsing module is of the same type, parse whether the highlight elements and their order of the highlight contents of the two are consistent; if one highlight element is different, mark the unpopular short video as a preferred highlight element short video of the split popular short video; if the highlight elements are completely consistent but the order of the highlight elements is not completely consistent, mark the unpopular short video as a short video with similar highlight elements of the split popular short video; if multiple highlight elements are inconsistent, check whether the highlight contents of the two exist in the same highlight metadata; if so, mark the unpopular short video as an extended short video of the recommended short video; if not, parse whether the similarity of the highlight clips of the two is greater than a specified similarity threshold; if so, mark the unpopular short video as a variant short video of the recommended short video; The third parsing module is used to, when the parsing result of the first parsing module is inconsistent in type, parse whether the highlight elements of the highlight contents of the two have one more highlight element or one less highlight element and the other highlight elements are completely consistent; if so, mark the unpopular short video as a short video with more highlight elements or fewer highlight elements of the split popular short video; if not, check whether the highlight contents of the two exist in the same highlight metadata; if so, mark the unpopular short video as an extended short video of the split popular short video; if not, parse whether the similarity of the highlight clips of the two is greater than a specified similarity threshold; if so, mark the unpopular short video as a variant short video of the split popular short video.

6. The system according to claim 5, characterized in that The comparison and selection module specifically includes: A comparison submodule, for comparing, for the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos with the highlight segments split from the popular short videos in the popular short video list; A first selection submodule is used to select a popular short video having the largest number of highlight segments compared with the highlight segments split from the unpopular short video as a recommended short video for the unpopular short video; The comparison and selection module specifically includes: A comparison submodule, for comparing, for the unpopular short videos in the unpopular short video list, the highlight segments split from the unpopular short videos with the highlight segments split from the popular short videos in the popular short video list; The second selection submodule is used to select a preset number of popular short videos as the hot short videos to be split from the unpopular short video in descending order of the number of highlight clips compared with the unpopular short video; A calculation submodule, used for outputting the similarity between the unpopular short video and each of the hot short videos to be split according to the number of the compared highlight segments and the number of the highlight segments split from the hot short videos to be split; The third selection submodule is used to select the hot short video to be split with the highest similarity as the recommended short video for the unpopular short video.

7. The system according to claim 6, characterized in that The comparison submodule is specifically used to compare the highlight segments of unpopular short videos in the unpopular short video list, whose number of highlight segments is greater than or equal to a first threshold, with the highlight segments of popular short videos in the popular short video list, whose number of highlight segments is greater than or equal to a second threshold; The calculation submodule is specifically used to calculate the ratio of the number of compared highlight clips to the number of highlight clips split from the hot short video to be split, and use the ratio as the similarity between the unpopular short video and the hot short video to be split.

8. The system according to claim 7, characterized in that The classification module specifically includes: A grading submodule, for grading the highlight metadata of each highlight short video in the highlight short video list, and outputting the highlight attribute information and the short video highlight content of each highlight metadata; The short video comparison submodule is used to compare the short video collection content of each collection metadata with the collection content of the hot short videos or the image attributes of the collection content in the corresponding collection attributes in the specified short video database, and output the hot short videos and the unpopular short videos; The first splicing submodule is used to splice the hot short videos and their split highlight clips that are consistent among all the hot short videos, and form a hot short video list with all the hot short videos and their split highlight clips output after the splicing operation; The second splicing submodule is used to splice the consistent unpopular short videos and their split highlight clips among all the unpopular short videos, and form an unpopular short video list with all the unpopular short videos and their split highlight clips output after the splicing operation.

Images