Topic identification methods, devices, electronic equipment and storage media

Abstract

This disclosure provides a topic identification method, apparatus, electronic device, and storage medium. The topic identification method includes: acquiring multiple video topics and videos for each video topic, wherein the videos for each video topic include published videos belonging to the video topic; acquiring video elements of each video in at least one content dimension; determining overlap index information of the multiple video topics based on the topic text of each video topic and the corresponding video elements, wherein the overlap index information is used to characterize the degree of overlap between pairs of video topics in the multiple video topics; and determining similar video topics for the same event among the multiple video topics based on the overlap index information.

Description

Technical Field

[0001] This disclosure generally relates to the field of data processing technology, and more specifically, to a topic identification method, apparatus, electronic device, and storage medium. Background Technology

[0002] As the video (e.g., short video) industry continues to develop, video consumption and creation have become part of people's daily lives. Video topics, as an important element of video content, play a significant role in both the production and consumption of content. For example, they are crucial for the recommendation ecosystem within video platforms. At the same time, topic tag pages provide a convenient path for content aggregation, satisfying users' demands for high-frequency consumption of certain types of content.

[0003] However, due to the fact that most video topics are freely edited by users, there are often many topics describing the same event or gameplay, which is not conducive to the accumulation of works under the topic and subsequent dissemination. Summary of the Invention

[0004] Exemplary embodiments of this disclosure provide a topic identification method, apparatus, electronic device, and storage medium capable of accurately identifying multiple similar video topics related to the same event.

[0005] According to a first aspect of the present disclosure, a topic identification method is provided, the topic identification method comprising: acquiring a plurality of video topics and videos for each video topic, wherein the videos for each video topic include published videos belonging to the video topic; acquiring video elements of each video in at least one content dimension; determining overlap index information of the plurality of video topics based on the topic text of each video topic and the video elements of the corresponding videos, wherein the overlap index information is used to characterize the degree of overlap between pairs of video topics in the plurality of video topics; and determining similar video topics for the same event among the plurality of video topics based on the overlap index information.

[0006] Optionally, the step of determining the overlap index information of the multiple video topics based on the topic text of each video topic and the video elements of the corresponding video includes: determining the text dimension overlap between the first video topic and the second video topic based on the topic text of the first video topic and the text of the corresponding video, and the topic text of the second video topic and the text of the corresponding video, wherein the first video topic and the second video topic are any two video topics among the multiple video topics; determining the audio dimension overlap between the first video topic and the second video topic based on the audio information of the video of the first video topic and the audio information of the video of the second video topic; determining the work dimension overlap between the first video topic and the second video topic based on the number of videos that simultaneously belong to the first video topic and the second video topic, the total number of videos of the first video topic, and the total number of videos of the second video topic; and obtaining the overlap index information of the multiple video topics based on at least one of the text dimension overlap, audio dimension overlap, and work dimension overlap between the first video topic and the second video topic.

[0007] Optionally, the step of determining the text dimension overlap between the first video topic and the second video topic based on the topic text of the first video topic and the corresponding video text, and the topic text of the second video topic and the corresponding video text includes: determining a first overlap between the topic text of the first video topic and the topic text of the second video topic; determining a second overlap between the video text of the first video topic and the video text of the second video topic; and determining the text dimension overlap between the first video topic and the second video topic based on the first overlap and the second overlap.

[0008] Optionally, the step of determining the audio dimension overlap between the first video topic and the second video topic based on the audio information of the videos of the first video topic and the audio information of the videos of the second video topic includes: identifying a first number of videos in the first video topic that use the same music as videos of the second video topic, and determining a first ratio between the first number and the total number of videos in the first video topic; identifying a second number of videos in the second video topic that use the same music as videos of the first video topic, and determining a second ratio between the second number and the total number of videos in the second video topic; and determining the audio dimension overlap between the first video topic and the second video topic based on the first ratio and the second ratio.

[0009] Optionally, the step of determining the overlap of the work dimension between the first video topic and the second video topic based on the number of videos simultaneously belonging to the first video topic and the second video topic, the total number of videos in the first video topic, and the total number of videos in the second video topic includes: identifying a third number of videos in the first video topic that also belong to the second video topic, and determining a third ratio between the third number and the total number of videos in the first video topic; identifying a fourth number of videos in the second video topic that also belong to the first video topic, and determining a fourth ratio between the fourth number and the total number of videos in the second video topic; and determining the overlap of the work dimension between the first video topic and the second video topic based on the third ratio and the fourth ratio.

[0010] Optionally, when the overlap index information includes text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap, the step of determining similar video topics for the same event among the multiple video topics based on the overlap index information includes: determining that the first video topic and the second video topic are similar video topics for the same event if the weighted sum of the text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap between the first video topic and the second video topic is greater than a first preset threshold; or, determining that the first video topic and the second video topic are similar video topics for the same event if the maximum value among the text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap between the first video topic and the second video topic is greater than a second preset threshold.

[0011] Optionally, the step of determining the first degree of overlap between the topic text of the first video topic and the topic text of the second video topic includes: determining the overlapping text portions of the topic text of the first video topic and the topic text of the second video topic, and determining the first degree of overlap based on the proportion of the overlapping text portions to the topic text of the first video topic and the proportion of the overlapping text portions to the topic text of the second video topic; and / or, determining the first degree of overlap based on the similarity between the feature vectors of the topic text of the first video topic and the feature vectors of the topic text of the second video topic; wherein, the step of determining the second degree of overlap between the video text of the first video topic and the video text of the second video topic includes: determining the overlapping text portions of the video text of the first video topic and the video text of the second video topic, and determining the second degree of overlap based on the proportion of the overlapping text portions to the text of the video text of the first video topic and the proportion of the overlapping text portions to the text of the video text of the second video topic; and / or, determining the second degree of overlap based on the similarity between the feature vectors of the video text of the first video topic and the feature vectors of the video text of the second video topic.

[0012] Optionally, the topic identification method further includes: displaying videos belonging to the video topic and videos belonging to similar video topics that address the same event within the tab of each video topic; and / or, for each video, recommending the video to the user based on the video topic to which the video belongs and similar video topics that address the same event.

[0013] According to a second aspect of the present disclosure, a topic identification device is provided, the topic identification device comprising: an acquisition unit configured to acquire a plurality of video topics and videos for each video topic, wherein the videos for each video topic include published videos belonging to the video topic; a video element acquisition unit configured to acquire video elements of each of the videos in at least one content dimension; an information determination unit configured to determine overlap index information of the plurality of video topics based on the topic text of each of the video topics and the video elements of the corresponding videos, wherein the overlap index information is used to characterize the degree of overlap between pairs of video topics among the plurality of video topics; and a determination unit configured to determine similar video topics among the plurality of video topics for the same event based on the overlap index information.

[0014] Optionally, the information determining unit is configured to: determine the text dimension overlap between the first video topic and the second video topic based on the topic text of the first video topic and the text of the corresponding video, and the topic text of the second video topic and the text of the corresponding video, wherein the first video topic and the second video topic are any two video topics among the plurality of video topics; determine the audio dimension overlap between the first video topic and the second video topic based on the audio information of the video of the first video topic and the audio information of the video of the second video topic; determine the work dimension overlap between the first video topic and the second video topic based on the number of videos that simultaneously belong to the first video topic and the second video topic, the total number of videos of the first video topic, and the total number of videos of the second video topic; and obtain the overlap index information of the plurality of video topics based on at least one of the text dimension overlap, audio dimension overlap, and work dimension overlap between the first video topic and the second video topic.

[0015] Optionally, the information determining unit is configured to: determine a first degree of overlap between the topic text of the first video topic and the topic text of the second video topic; determine a second degree of overlap between the text of the video of the first video topic and the text of the video of the second video topic; and determine the text dimension overlap between the first video topic and the second video topic based on the first degree of overlap and the second degree of overlap.

[0016] Optionally, the information determining unit is configured to: identify a first number of videos in the first video topic that use the same music as videos in the second video topic, and determine a first ratio between the first number and the total number of videos in the first video topic; identify a second number of videos in the second video topic that use the same music as videos in the first video topic, and determine a second ratio between the second number and the total number of videos in the second video topic; and determine the audio dimension overlap between the first video topic and the second video topic based on the first ratio and the second ratio.

[0017] Optionally, the information determining unit is configured to: identify a third number of videos that simultaneously belong to the second video topic among the videos of the first video topic, and determine a third ratio between the third number and the total number of videos of the first video topic; identify a fourth number of videos that simultaneously belong to the first video topic among the videos of the second video topic, and determine a fourth ratio between the fourth number and the total number of videos of the second video topic; and determine the degree of overlap in the work dimensions between the first video topic and the second video topic based on the third ratio and the fourth ratio.

[0018] Optionally, when the overlap index information includes text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap, the determining unit is configured to: determine that the first video topic and the second video topic are similar video topics concerning the same event when the weighted sum of the text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap between the first video topic and the second video topic is greater than a first preset threshold; or, determine that the first video topic and the second video topic are similar video topics concerning the same event when the maximum value among the text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap between the first video topic and the second video topic is greater than a second preset threshold.

[0019] Optionally, the information determining unit is configured to: determine the text portions that overlap between the topic text of the first video topic and the topic text of the second video topic, and determine a first degree of overlap based on the proportion of the overlapping text portions to the topic text of the first video topic and the proportion of the overlapping text portions to the topic text of the second video topic; and / or, determine the first degree of overlap based on the similarity between the feature vectors of the topic text of the first video topic and the feature vectors of the topic text of the second video topic; the information determining unit is configured to: determine the text portions that overlap between the video text of the first video topic and the video text of the second video topic, and determine a second degree of overlap based on the proportion of the overlapping text portions to the video text of the first video topic and the proportion of the overlapping text portions to the video text of the second video topic; and / or, determine the second degree of overlap based on the similarity between the feature vectors of the video text of the first video topic and the feature vectors of the video text of the second video topic.

[0020] Optionally, the topic identification device further includes: a tab providing unit, configured to display videos belonging to the video topic and videos belonging to similar video topics that address the same event as the video topic within the tab of each video topic; and / or, a recommendation unit, configured to recommend the video to the user for each video based on the video topic to which the video belongs and similar video topics that address the same event as the video topic.

[0021] According to a third aspect of the present disclosure, an electronic device is provided, comprising: at least one processor; and at least one memory storing computer-executable instructions, wherein the computer-executable instructions, when executed by the at least one processor, cause the at least one processor to perform the topic identification method as described above.

[0022] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided, wherein when instructions in the computer-readable storage medium are executed by at least one processor, the at least one processor causes the at least one processor to perform the topic identification method as described above.

[0023] According to a fifth aspect of the present disclosure, a computer program product is provided, including computer instructions that, when executed by at least one processor, implement the topic identification method as described above.

[0024] The topic identification method, apparatus, electronic device, and storage medium according to exemplary embodiments of this disclosure identify multiple similar video topics for the same event not only based on the topic text of the video topic itself, but also based on the information of the video under the video topic, thereby improving the accuracy of video topic identification.

[0025] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0026] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure, and are not intended to unduly limit this disclosure.

[0027] Figure 1 A flowchart illustrating a topic identification method according to an exemplary embodiment of this disclosure is provided.

[0028] Figure 2 A flowchart illustrating a method for determining overlap index information of multiple video topics according to an exemplary embodiment of the present disclosure;

[0029] Figure 3 A flowchart illustrating a method for determining the textual dimension overlap of any two video topics according to an exemplary embodiment of this disclosure;

[0030] Figure 4 A flowchart illustrating a method for determining the audio dimension overlap of any two video topics according to an exemplary embodiment of the present disclosure;

[0031] Figure 5 A flowchart illustrating a method for determining the overlap of work dimensions between any two video topics according to an exemplary embodiment of this disclosure;

[0032] Figure 6 An example of a topic identification method according to an exemplary embodiment of this disclosure is shown;

[0033] Figure 7 A structural block diagram of a topic identification device according to an exemplary embodiment of the present disclosure is shown;

[0034] Figure 8 A structural block diagram of an electronic device according to an exemplary embodiment of the present disclosure is shown. Detailed Implementation

[0035] To enable those skilled in the art to better understand the technical solutions of this disclosure, the technical solutions in the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings.

[0036] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0037] It should be noted that the phrase "at least one of several items" in this disclosure refers to three parallel cases: "any one of the several items", "a combination of any number of the several items", and "all of the several items". For example, "including at least one of A and B" includes the following three parallel cases: (1) including A; (2) including B; (3) including A and B. As another example, "performing at least one of step one and step two" indicates the following three parallel cases: (1) performing step one; (2) performing step two; (3) performing both step one and step two.

[0038] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for display, data used for analysis, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties.

[0039] Figure 1 A flowchart illustrating a topic identification method according to an exemplary embodiment of the present disclosure is shown.

[0040] Reference Figure 1 In step S101, multiple video topics and videos for each video topic are obtained.

[0041] Each video topic includes videos that have been published and belong to that video topic.

[0042] As an example, the aforementioned video topic can be a short video topic, and the aforementioned video type can be a short video. It should be understood that other types of video topics and videos are also possible, and this disclosure does not impose any restrictions on them.

[0043] As an example, videos published within the most recent preset time period that are tagged with at least one video topic can be obtained; then, the video topics to which the obtained videos belong are taken as the multiple video topics; and based on the obtained videos, videos for each video topic are determined. For example, for each video topic, a certain number of videos belonging to that video topic can be selected as the videos for that video topic for subsequent execution of step S102. It should be understood that when the same video is tagged with multiple video topics, the video belongs to all of these video topics simultaneously, that is, under each of these multiple video topics.

[0044] For example, users can tag one or more video topics (i.e., the video topics they belong to) based on the content of a video they are about to publish, so that after the video is published, it can be associated with the tagged video topics. For example, the video can be displayed on the tab page of its respective video topic. For example, the video can be recommended to other users based on its respective video topic.

[0045] In step S102, video elements of each video in at least one content dimension are obtained.

[0046] As an example, the at least one content dimension may include, but is not limited to, at least one of the following: text dimension, audio dimension. Correspondingly, video elements may include, but are not limited to, at least one of the following: text, audio information. It should be understood that other content dimensions and video elements may also be included, and this disclosure does not limit this.

[0047] As an example, the audio information of a video may include information about the music used in the video. For instance, the audio information of a video may include identification or characteristic information of the music used in the video (e.g., background music). It should be understood that other types of audio information may also be included, and this disclosure is not limiting in this regard.

[0048] As an example, the audio information of a video may include information extracted from the video's audio using audio fingerprinting technology. For instance, audio fingerprinting technology uses specific algorithms to extract unique digital features from an audio segment as identifiers, which can be used to identify massive amounts of sound samples or track the location of samples in a database. In original soundtracks (user-edited works), it can identify which piece of music is used as the background music.

[0049] As an example, the text of a video may include, but is not limited to, at least one of the following: video script, text identified from images in the video, and text converted from the audio of the video. It should be understood that other types of text may also be included, and this disclosure is not limiting in this regard.

[0050] As an example, OCR (optical character recognition) can be used to identify text from images in a video. For instance, OCR can be performed on video frames such as cover frames and keyframes to extract text from the video content.

[0051] As an example, ASR (Automatic Speech Recognition) technology can be used to convert the audio of a video into text in order to extract the audio content of the video.

[0052] As an example, video copy can include the video title, descriptive text, etc.

[0053] In step S103, the overlap index information of the multiple video topics is determined based on the topic text of each video topic and the video elements of the corresponding video.

[0054] The overlap index information is used to characterize the degree of overlap between any two video topics among the multiple video topics.

[0055] As an example, overlap metrics may include, but are not limited to, at least one of the following: text-level overlap, audio-level overlap, and work-level overlap.

[0056] Textual overlap is used to characterize the degree of text overlap between two video topics. The text here includes the text of the video topic itself and the text of the videos under the video topic.

[0057] Audio overlap is used to characterize the degree of audio overlap between two video topics. Here, audio includes the audio information of videos under the same video topic.

[0058] Work dimension overlap is used to characterize the degree of overlap between video topics in pairs of videos.

[0059] Step S103 can be implemented in various appropriate ways, which will be discussed below. Figure 2 An exemplary embodiment of step S103 will be described below.

[0060] In step S104, based on the overlap index information, similar video topics concerning the same event are determined among the multiple video topics.

[0061] According to an exemplary embodiment of this disclosure, based on the association between video topics and videos under video topics, information beyond the text of the video topic itself (i.e., multimodal video element information of videos under video topics, etc.) is extended to identify similar video topics for the same event, thereby improving the accuracy of video topic identification.

[0062] Furthermore, as an example, the topic identification method according to an exemplary embodiment of this disclosure may further include: displaying videos belonging to that video topic and videos belonging to similar video topics that address the same event within each video topic's tab. Thus, users can see more desired videos within the video topic tabs.

[0063] Furthermore, as an example, the topic identification method according to an exemplary embodiment of this disclosure may further include: for each video, recommending the video to the user based on the video topic to which the video belongs and similar video topics that address the same event as the video topic. Thus, it is possible to recommend videos that better suit the user's needs.

[0064] The topic identification method according to the exemplary embodiments of this disclosure can accurately identify multiple video topics under the same event / gameplay, thereby understanding the true popularity of a certain event / gameplay and facilitating the selection of high-quality topics for operation. From a user experience perspective, by accurately identifying multiple video topics under the same event / gameplay, users can be recommended higher-quality video topics and videos that better meet their needs, and users will see more videos within the video topic tab.

[0065] Figure 2 A flowchart illustrating a method for determining overlap index information of multiple video topics according to an exemplary embodiment of the present disclosure is shown.

[0066] Reference Figure 2 In step S201, based on the topic text of the first video topic and the text of the corresponding video, and the topic text of the second video topic and the text of the corresponding video, the text dimension overlap between the first video topic and the second video topic is determined.

[0067] The first video topic and the second video topic are any two video topics from the plurality of video topics.

[0068] The following will combine Figure 3 An exemplary embodiment of step S201 will be described below. It should be understood that other appropriate methods may also be used to determine the text dimension overlap.

[0069] In step S202, based on the audio information of the video of the first video topic and the audio information of the video of the second video topic, the audio dimension overlap between the first video topic and the second video topic is determined.

[0070] The following will combine Figure 4 An exemplary embodiment of step S202 will be described below. It should be understood that other suitable methods may also be used to determine the audio dimension overlap.

[0071] In step S203, the overlap of work dimensions between the first video topic and the second video topic is determined based on the number of videos that simultaneously belong to the first video topic and the second video topic, the total number of videos in the first video topic, and the total number of videos in the second video topic.

[0072] The following will combine Figure 5 An exemplary embodiment of step S203 will be described below. It should be understood that other appropriate methods may also be used to determine the overlap of work dimensions.

[0073] In step S204, the overlap index information of the multiple video topics is obtained based on at least one of the text dimension overlap, audio dimension overlap, and work dimension overlap between the first video topic and the second video topic.

[0074] According to exemplary embodiments of this disclosure, it is possible to reasonably and effectively determine the overlap index information of multiple video topics.

[0075] As an example, step S104 may include: determining whether the first video topic and the second video topic are similar video topics for the same event based on the overlap of text dimension, audio dimension, and work dimension between the first video topic and the second video topic.

[0076] As an example, if the weighted sum of the text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap between the first and second video topics exceeds a first preset threshold, then the first and second video topics can be determined to be similar video topics concerning the same event. For instance, the weight of each item can be 1 / 3, meaning that if the average of the text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap between the first and second video topics exceeds the first preset threshold, then the first and second video topics can be determined to be similar video topics concerning the same event.

[0077] As another example, if the maximum value among the text dimension overlap, audio dimension overlap, and work dimension overlap between the first video topic and the second video topic is greater than a second preset threshold, it can be determined that the first video topic and the second video topic are similar video topics for the same event.

[0078] As an example, if video topic A and video topic B are similar video topics about the same event, and video topic A and video topic C are similar video topics about the same event, then video topic A, video topic B, and video topic C can be determined to be similar video topics about the same event.

[0079] According to exemplary embodiments of this disclosure, by utilizing multimodal information of videos under a video topic, similar video topics for the same event can be conveniently and effectively identified.

[0080] Figure 3 A flowchart illustrating a method for determining the textual dimension overlap of any two video topics according to an exemplary embodiment of this disclosure is provided. The arbitrary two video topics include a first video topic and a second video topic.

[0081] Reference Figure 3 In step S301, the first degree of overlap between the topic text of the first video topic and the topic text of the second video topic is determined.

[0082] As an example, step S301 may include: determining the text portions that overlap between the topic text of the first video topic and the topic text of the second video topic; determining a first degree of overlap based on the proportion of the overlapping text portions to the topic text of the first video topic and the proportion of the overlapping text portions to the topic text of the second video topic; and / or determining the first degree of overlap based on the similarity between the feature vectors of the topic text of the first video topic and the feature vectors of the topic text of the second video topic.

[0083] It should be understood that the degree of overlap between the topic text of the first video topic and the topic text of the second video topic can also be determined through other appropriate methods.

[0084] For example, the first degree of overlap can be the weighted sum or the maximum value of the proportion of the overlapping text portion to the topic text of the first video topic and the proportion of the overlapping text portion to the topic text of the second video topic.

[0085] For example, the similarity between the feature vectors of the topic text of the first video topic and the feature vectors of the topic text of the second video topic can be directly used as the first degree of overlap. For example, the feature vector of the topic text can be the embedding vector of the topic text. As an example, the type of similarity between feature vectors can be cosine similarity.

[0086] For example, the overlap degrees determined by the above two methods can be weighted and summed to obtain the final first overlap degree.

[0087] As an example, overlapping text portions can be identified through text segmentation, part-of-speech analysis, semantic analysis, and feature extraction. For instance, overlapping text portions may include completely identical text portions or text portions with similar content.

[0088] In step S302, a second degree of overlap is determined between the text of the video for the first video topic and the text of the video for the second video topic.

[0089] As an example, step S302 may include: determining the text portions that overlap between the text of a video under a first video topic and the text of a video under a second video topic; determining a second degree of overlap based on the proportion of the overlapping text portions to the text of the video under the first video topic and the proportion of the overlapping text portions to the text of the video under the second video topic; and / or determining the second degree of overlap based on the similarity between the feature vectors of the text of the video under the first video topic and the feature vectors of the text of the video under the second video topic.

[0090] In one embodiment, all types of text from all videos under a first video topic can be aggregated together as first text, and all types of text from all videos under a second video topic can be aggregated together as second text. Then, the first text and the second text are compared to determine a second degree of overlap.

[0091] In another embodiment, the text of each type of video under each video topic can be aggregated together, and a comparison between the first video topic and the second video topic can be performed on this type of text. For example, for the text of the video script, the comparison between the first video topic and the second video topic is used to obtain the script overlap; for the text of the text identified from the video image, the comparison between the first video topic and the second video topic is used to obtain the image text overlap; for the text of the text converted from the video speech, the comparison between the first video topic and the second video topic is used to obtain the speech text overlap; then, the weighted sum or the maximum value of the above three overlap values ​​is taken as the final second overlap.

[0092] As an example, the feature vector of text can be the text's embedding vector. As another example, the type of similarity between feature vectors can be cosine similarity.

[0093] Furthermore, the second degree of overlap can be determined based on the above example of the method for determining the first degree of overlap, which will not be elaborated here.

[0094] According to exemplary embodiments of this disclosure, text overlap can be determined conveniently and effectively.

[0095] In step S303, the text dimension overlap between the first video topic and the second video topic is determined based on the first overlap and the second overlap.

[0096] As an example, the weighted sum or maximum value of the first and second overlap can be used as the text dimension overlap between the first and second video topics.

[0097] According to an exemplary embodiment of this disclosure, by utilizing the multimodal information of videos under a video topic, the text dimension overlap of any two video topics can be accurately and comprehensively determined.

[0098] Figure 4 A flowchart illustrating a method for determining the audio dimension overlap of any two video topics according to an exemplary embodiment of this disclosure is provided. The arbitrary two video topics include a first video topic and a second video topic.

[0099] Reference Figure 4 In step S401, the first number of videos under the first video topic that use the same music as videos under the second video topic is identified, and a first ratio between the first number and the total number of videos under the first video topic is determined.

[0100] As an example, for each video under the first video topic, it can be identified whether there is any video under the second video topic that uses the same music. If so, it is identified as a video that uses the same music as a video under the second video topic.

[0101] In step S402, a second number of videos under the second video topic that use the same music as videos under the first video topic are identified, and a second ratio between the second number and the total number of videos under the second video topic is determined.

[0102] As an example, for each video under the second video topic, it can be identified whether there is any video under the first video topic that uses the same music. If so, it is identified as a video that uses the same music as a video under the first video topic.

[0103] In step S403, based on the first ratio and the second ratio, the audio dimension overlap between the first video topic and the second video topic is determined, that is, the music overlap is obtained.

[0104] As an example, the weighted sum or maximum value of the first and second ratios can be used as the audio dimension overlap between the first and second video topics.

[0105] According to an exemplary embodiment of this disclosure, by utilizing the multimodal information of videos under a video topic, the audio dimension overlap of any two video topics can be accurately and comprehensively determined.

[0106] Figure 5 A flowchart illustrating a method for determining the overlap of work dimensions between any two video topics according to an exemplary embodiment of this disclosure is provided. The arbitrary two video topics include a first video topic and a second video topic.

[0107] Reference Figure 5 In step S501, a third number of videos that also belong to the second video topic are identified among the videos under the first video topic, and a third ratio between the third number and the total number of videos under the first video topic is determined.

[0108] In step S502, the fourth number of videos that also belong to the first video topic are identified among the videos under the second video topic, and the fourth ratio between the fourth number and the total number of videos under the second video topic is determined.

[0109] In step S503, the overlap of work dimensions between the first video topic and the second video topic is determined based on the third ratio and the fourth ratio.

[0110] As an example, the weighted sum or maximum value of the third and fourth ratios can be used as the overlap of the work dimensions between the first and second video topics.

[0111] According to exemplary embodiments of this disclosure, the degree of overlap in the work dimensions of any two video topics can be accurately determined.

[0112] Figure 6 An example of a topic identification method according to an exemplary embodiment of this disclosure is shown.

[0113] like Figure 6 As shown, multiple video topics are identified based on user-posted videos within a certain period. These video topics are aggregated across the topic text, audio fingerprint, and video text dimensions. Specifically, through data statistics and mining, overlap is calculated across three dimensions: text, music, and content. The similarity between any two video topics is then aggregated, and a threshold is used to determine whether any two video topics are similar (i.e., video topics related to the same event / gameplay). The final judgment result is then output.

[0114] Text overlap calculation: For the text of videos under a video topic (e.g., including the text of each video, the text identified from the images of each video, and the text converted from the audio of each video), as well as the text of the topic itself, the text overlap can be calculated in the following two ways: 1. Calculate the overlap by calculating the ratio of overlapping / similar text to the overall text; 2. Calculate the cosine similarity by using the text embedding.

[0115] Music overlap calculation: By using audio fingerprint technology, we can identify which music is the original soundtrack of a video under a video topic. By calculating the ratio of works using the same song under two topics to all works, we can obtain the music overlap.

[0116] Work overlap calculation: For two video topics, the overlap is calculated by measuring the number of works that simultaneously mention both topics and their proportion to all works under each topic.

[0117] Final overlap calculation and threshold determination: The final overlap can be calculated in two ways. One is to take the average overlap from all sources, and the other is to take the maximum overlap from all sources and then use a threshold (e.g., a overlap greater than 80% can be considered as similar topics about the same event) to determine whether they are similar topics about the same event. For example, video topic 1 and video topic 2 are ultimately determined to be similar video topics about the same event.

[0118] Figure 7 A structural block diagram of a topic recognition device according to an exemplary embodiment of the present disclosure is shown.

[0119] like Figure 7 As shown, the topic identification device 10 according to an exemplary embodiment of the present disclosure includes: an acquisition unit 101, a video element acquisition unit 102, an information determination unit 103, and a determination unit 104.

[0120] Specifically, the acquisition unit 101 is configured to acquire multiple video topics and videos for each video topic, wherein the videos for each video topic include published videos that belong to the video topic.

[0121] The video element acquisition unit 102 is configured to acquire video elements of each of the videos in at least one content dimension.

[0122] The information determination unit 103 is configured to determine overlap index information of the plurality of video topics based on the topic text of each video topic and the video elements of the corresponding video, wherein the overlap index information is used to characterize the degree of overlap between any two video topics among the plurality of video topics.

[0123] The determining unit 104 is configured to determine similar video topics for the same event among the multiple video topics based on the overlap index information.

[0124] As an example, the information determination unit 103 may be configured to: determine the text dimension overlap between the first video topic and the second video topic based on the topic text of the first video topic and the text of the corresponding video, and the topic text of the second video topic and the text of the corresponding video, wherein the first video topic and the second video topic are any two video topics among the plurality of video topics; determine the audio dimension overlap between the first video topic and the second video topic based on the audio information of the video of the first video topic and the audio information of the video of the second video topic; determine the work dimension overlap between the first video topic and the second video topic based on the number of videos that simultaneously belong to the first video topic and the second video topic, the total number of videos of the first video topic, and the total number of videos of the second video topic; and obtain the overlap index information of the plurality of video topics based on at least one of the text dimension overlap, audio dimension overlap, and work dimension overlap between the first video topic and the second video topic.

[0125] As an example, the information determination unit 103 may be configured to: determine a first degree of overlap between the topic text of the first video topic and the topic text of the second video topic; determine a second degree of overlap between the text of the video of the first video topic and the text of the video of the second video topic; and determine the text dimension overlap between the first video topic and the second video topic based on the first degree of overlap and the second degree of overlap.

[0126] As an example, the information determination unit 103 may be configured to: identify a first number of videos in the first video topic that use the same music as videos in the second video topic, and determine a first ratio between the first number and the total number of videos in the first video topic; identify a second number of videos in the second video topic that use the same music as videos in the first video topic, and determine a second ratio between the second number and the total number of videos in the second video topic; and determine the audio dimension overlap between the first video topic and the second video topic based on the first ratio and the second ratio.

[0127] As an example, the information determination unit 103 may be configured to: identify a third number of videos that simultaneously belong to the second video topic among the videos of the first video topic, and determine a third ratio between the third number and the total number of videos of the first video topic; identify a fourth number of videos that simultaneously belong to the first video topic among the videos of the second video topic, and determine a fourth ratio between the fourth number and the total number of videos of the second video topic; and determine the degree of overlap in the work dimension between the first video topic and the second video topic based on the third ratio and the fourth ratio.

[0128] As an example, when the overlap index information includes text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap, the determining unit 104 can be configured to: determine that the first video topic and the second video topic are similar video topics for the same event when the weighted sum of the text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap between the first video topic and the second video topic is greater than a first preset threshold; or, determine that the first video topic and the second video topic are similar video topics for the same event when the maximum value among the text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap between the first video topic and the second video topic is greater than a second preset threshold.

[0129] As an example, the information determination unit 103 may be configured to: determine the text portions that overlap between the topic text of the first video topic and the topic text of the second video topic; determine a first degree of overlap based on the proportion of the overlapping text portions to the topic text of the first video topic and the proportion of the overlapping text portions to the topic text of the second video topic; and / or determine the first degree of overlap based on the similarity between the feature vectors of the topic text of the first video topic and the feature vectors of the topic text of the second video topic.

[0130] As an example, the information determination unit 103 may be configured to: determine the text portions that overlap between the text of the video of the first video topic and the text of the video of the second video topic; determine a second degree of overlap based on the proportion of the overlapping text portions to the text of the video of the first video topic and the proportion of the overlapping text portions to the text of the video of the second video topic; and / or determine the second degree of overlap based on the similarity between the feature vectors of the text of the video of the first video topic and the feature vectors of the text of the video of the second video topic.

[0131] As an example, the topic identification device 10 may further include: a tab providing unit (not shown) and / or a recommendation unit (not shown), wherein the tab providing unit is configured to display videos belonging to the video topic and videos belonging to similar video topics that address the same event as the video topic within the tab of each video topic; and the recommendation unit is configured to recommend the video to the user for each video based on the video topic to which the video belongs and similar video topics that address the same event as the video topic.

[0132] Regarding the topic recognition device 10 in the above embodiments, the specific way in which each unit performs operations has been described in detail in the embodiments related to the method, and will not be elaborated here.

[0133] Furthermore, it should be understood that the various units in the topic identification apparatus 10 according to the exemplary embodiments of this disclosure may be implemented as hardware components and / or software components. Those skilled in the art, based on the processes performed by the defined various units, may implement the various units, for example, using a field-programmable gate array (FPGA) or an application-specific integrated circuit (ASIC).

[0134] Figure 8 A structural block diagram of an electronic device according to an exemplary embodiment of the present disclosure is shown. (Refer to...) Figure 8 The electronic device 20 includes at least one memory 201 and at least one processor 202. The at least one memory 201 stores a set of computer-executable instructions. When the set of computer-executable instructions is executed by the at least one processor 202, the topic identification method as described in the exemplary embodiments above is executed.

[0135] As an example, electronic device 20 may be a PC, tablet, personal digital assistant, smartphone, or other device capable of executing the aforementioned set of instructions. Here, electronic device 20 is not necessarily a single electronic device; it may be any collection of devices or circuits capable of executing the aforementioned instructions (or instruction sets) individually or in combination. Electronic device 20 may also be part of an integrated control system or system manager, or may be configured to interconnect with a portable electronic device locally or remotely (e.g., via wireless transmission) through an interface.

[0136] In electronic device 20, processor 202 may include a central processing unit (CPU), a graphics processing unit (GPU), a programmable logic device, a dedicated processor system, a microcontroller, or a microprocessor. By way of example and not limitation, processor 202 may also include analog processors, digital processors, microprocessors, multi-core processors, processor arrays, network processors, etc.

[0137] The processor 202 can execute instructions or code stored in the memory 201, which can also store data. Instructions and data can also be sent and received over a network via a network interface device, which can employ any known transmission protocol.

[0138] The memory 201 may be integrated with the processor 202, for example, by placing RAM or flash memory within an integrated circuit microprocessor. Alternatively, the memory 201 may include a separate device, such as an external disk drive, a storage array, or other storage device that can be used by any database system. The memory 201 and the processor 202 may be operatively coupled, or may communicate with each other, for example, via I / O ports, network connections, etc., enabling the processor 202 to read files stored in the memory.

[0139] In addition, electronic device 20 may also include a video display (such as a liquid crystal display) and a user interaction interface (such as a keyboard, mouse, touch input device, etc.). All components of electronic device 20 may be interconnected via a bus and / or network.

[0140] According to exemplary embodiments of this disclosure, a computer-readable storage medium storing instructions may also be provided, wherein when the instructions are executed by at least one processor, they cause at least one processor to perform the topic identification method as described in the exemplary embodiments above. Examples of computer-readable storage media herein include: read-only memory (ROM), random access programmable read-only memory (PROM), electrically erasable programmable read-only memory (EEPROM), random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), flash memory, non-volatile memory, CD-ROM, CD-R, CD+R, CD-RW, CD+RW, DVD-ROM, DVD-R, DVD+R, DVD-RW, DVD+RW, DVD-RAM, BD-ROM, BD-R, BD-R LTH, BD-RE, Blu-ray or optical disc storage, hard disk drive (HDD), solid-state drive (SSD), card storage (such as multimedia cards, secure digital (SD) cards, or ultra-fast digital (XD) cards), magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid-state drive, and any other device configured to store a computer program and any associated data, data files, and data structures in a non-transitory manner and to provide the computer program and any associated data, data files, and data structures to a processor or computer so that the processor or computer can execute the computer program. The computer program in the aforementioned computer-readable storage medium can run in an environment deployed in computer devices such as clients, hosts, agent devices, servers, etc. Furthermore, in one example, the computer program and any associated data, data files, and data structures are distributed across a networked computer system, such that the computer program and any associated data, data files, and data structures are stored, accessed, and executed in a distributed manner through one or more processors or computers.

[0141] According to exemplary embodiments of the present disclosure, a computer program product may also be provided, wherein the instructions in the computer program product are executable by at least one processor to perform the topic identification method as described in the exemplary embodiments above.

[0142] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.

[0143] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. A topic identification method, characterized in that, The topic identification method includes: Retrieve multiple video topics and videos for each video topic, wherein the videos for each video topic include published videos that belong to the video topic; Obtain the video elements of each video under at least one content dimension; Based on the topic text of each video topic and the corresponding video elements, the overlap index information of the multiple video topics is determined, wherein the overlap index information is used to characterize the degree of overlap between any two video topics among the multiple video topics. Based on the overlap index information, similar video topics concerning the same event are identified among the multiple video topics; The step of determining the overlap index information of the multiple video topics based on the topic text of each video topic and the corresponding video elements includes: Based on the topic text of the first video topic and the text of the corresponding video, and the topic text of the second video topic and the text of the corresponding video, the text dimension overlap between the first video topic and the second video topic is determined, wherein the first video topic and the second video topic are any two video topics among the plurality of video topics. Based on the audio information of the video of the first video topic and the audio information of the video of the second video topic, the degree of audio dimension overlap between the first video topic and the second video topic is determined. The overlap of work dimensions between the first video topic and the second video topic is determined based on the number of videos that simultaneously belong to the first video topic and the second video topic, the total number of videos in the first video topic, and the total number of videos in the second video topic. Based on the overlap in text dimension, audio dimension, and work dimension between the first video topic and the second video topic, overlap index information of the multiple video topics is obtained.

2. The topic identification method according to claim 1, characterized in that, The step of determining the text dimension overlap between the first video topic and the second video topic based on the topic text of the first video topic and the corresponding video text, and the topic text of the second video topic and the corresponding video text, includes: Determine the first degree of overlap between the topic text of the first video topic and the topic text of the second video topic; Determine the second degree of overlap between the text of the video on the first video topic and the text of the video on the second video topic; Based on the first overlap and the second overlap, the text dimension overlap between the first video topic and the second video topic is determined.

3. The topic identification method according to claim 1, characterized in that, The step of determining the audio dimension overlap between the first video topic and the second video topic based on the audio information of the video of the first video topic and the audio information of the video of the second video topic includes: Identify a first number of videos in the first video topic that use the same music as videos in the second video topic, and determine a first ratio between the first number and the total number of videos in the first video topic. Identify a second number of videos in the second video topic that use the same music as videos in the first video topic, and determine a second ratio between the second number and the total number of videos in the second video topic; Based on the first ratio and the second ratio, the audio dimension overlap between the first video topic and the second video topic is determined.

4. The topic identification method according to claim 1, characterized in that, The step of determining the overlap of work dimensions between the first video topic and the second video topic based on the number of videos simultaneously belonging to the first video topic and the second video topic, the total number of videos in the first video topic, and the total number of videos in the second video topic includes: Identify a third number of videos that simultaneously belong to the second video topic among the videos of the first video topic, and determine a third ratio between the third number and the total number of videos of the first video topic; Identify a fourth number of videos that simultaneously belong to the first video topic among the videos of the second video topic, and determine a fourth ratio between the fourth number and the total number of videos of the second video topic; Based on the third ratio and the fourth ratio, the overlap of work dimensions between the first video topic and the second video topic is determined.

5. The topic identification method according to claim 1, characterized in that, When the overlap index information includes text-dimensional overlap, audio-dimensional overlap, and work-dimensional overlap, the step of determining similar video topics concerning the same event among the multiple video topics based on the overlap index information includes: If the weighted sum of the text dimension overlap, audio dimension overlap, and work dimension overlap between the first video topic and the second video topic is greater than a first preset threshold, it is determined that the first video topic and the second video topic are similar video topics for the same event. Alternatively, if the maximum value among the text dimension overlap, audio dimension overlap, and work dimension overlap between the first video topic and the second video topic is greater than a second preset threshold, it is determined that the first video topic and the second video topic are similar video topics concerning the same event.

6. The topic identification method according to claim 2, characterized in that, The step of determining the first degree of overlap between the topic text of the first video topic and the topic text of the second video topic includes: Identify the overlapping text portions of the topic text of the first video topic and the topic text of the second video topic, and determine the first degree of overlap based on the proportion of the overlapping text portions to the topic text of the first video topic and the proportion of the overlapping text portions to the topic text of the second video topic; and / or, determine the first degree of overlap based on the similarity between the feature vectors of the topic text of the first video topic and the feature vectors of the topic text of the second video topic. The step of determining the second degree of overlap between the text of the video on the first video topic and the text of the video on the second video topic includes: Identify the text portions that overlap between the text of the video of the first video topic and the text of the video of the second video topic, and determine a second degree of overlap based on the proportion of the overlapping text portions to the text of the video of the first video topic and the proportion of the overlapping text portions to the text of the video of the second video topic; and / or, determine the second degree of overlap based on the similarity between the feature vectors of the text of the video of the first video topic and the feature vectors of the text of the video of the second video topic.

7. The topic identification method according to claim 1, characterized in that, The topic identification method also includes: Within each video topic's tab, videos belonging to that video topic and videos belonging to similar video topics that address the same event as that video topic are displayed; And / or, for each video, recommend the video to the user based on the video topic to which the video belongs and similar video topics that address the same event.

8. A topic recognition device, characterized in that, The topic identification device includes: The acquisition unit is configured to acquire multiple video topics and videos for each video topic, wherein the videos for each video topic include published videos that belong to the video topic; The video element acquisition unit is configured to acquire video elements of each of the videos in at least one content dimension; The information determination unit is configured to determine overlap index information of the multiple video topics based on the topic text of each video topic and the video elements of the corresponding video, wherein the overlap index information is used to characterize the degree of overlap between any two video topics among the multiple video topics. The determining unit is configured to determine similar video topics concerning the same event among the multiple video topics based on the overlap index information. The information determining unit is further configured to: determine the text dimension overlap between the first video topic and the second video topic based on the topic text and corresponding video text of the first video topic, and the topic text and corresponding video text of the second video topic, wherein the first video topic and the second video topic are any two of the plurality of video topics; determine the audio dimension overlap between the first video topic and the second video topic based on the audio information of the video of the first video topic and the audio information of the video of the second video topic; determine the work dimension overlap between the first video topic and the second video topic based on the number of videos that simultaneously belong to the first video topic and the second video topic, the total number of videos of the first video topic, and the total number of videos of the second video topic; and obtain the overlap index information of the plurality of video topics based on the text dimension overlap, audio dimension overlap, and work dimension overlap between the first video topic and the second video topic.

9. An electronic device, characterized in that, include: At least one processor; At least one memory that stores computer-executable instructions. Wherein, when the computer-executable instructions are executed by the at least one processor, the at least one processor causes the at least one processor to perform the topic identification method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, When the instructions in the computer-readable storage medium are executed by at least one processor, the at least one processor causes the at least one processor to perform the topic identification method as described in any one of claims 1 to 7.

Images