Methods for determining video tags, methods for displaying video list information, and related devices

Abstract

This invention provides a method for determining video tags, a method for displaying video list information, and related apparatus. The method includes: acquiring interactive text posted for a target video; clustering the interactive text based on the text feature similarity to obtain one or more interactive text clusters and clustering tags for each interactive text cluster; and determining topic tags for the target video based on the clustering tags of the interactive text clusters. This method can determine more intelligent, detailed, and diverse video tags.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method for determining video tags, a method for displaying video list information, and related devices. Background Technology

[0002] Currently, video tags are widely used in tasks such as video classification, recommendation, and retrieval. Therefore, the accuracy and diversity of video tags are crucial for video management and promotion.

[0003] However, current video tags are mainly generated based on actors, themes, etc., making it difficult to utilize personalized information spread by users and to automatically track user-generated topics related to TV series on a large scale, resulting in insufficient diversity of video tags. Summary of the Invention

[0004] The purpose of this invention is to provide a method for determining video tags, a method for displaying video list information, and related devices to enhance the diversity of video tags. The specific technical solution is as follows:

[0005] In a first aspect of this invention, a video tag determination method is provided, comprising:

[0006] Retrieve interactive text posted for the target video;

[0007] Based on the text feature similarity of the interactive texts, the interactive texts are clustered to obtain one or more interactive text clusters and clustering labels for each interactive text cluster;

[0008] Based on the clustering tags of the interactive text clusters, the topic tags of the target video are determined.

[0009] Optionally, the method further includes:

[0010] The text similarity between the topic tags determined for multiple target videos is determined, and the target videos corresponding to the topic tags whose text similarity is not less than the similarity threshold are aggregated to obtain a video recommendation list.

[0011] Optionally, the step of clustering the interactive text based on the text feature similarity of the interactive text includes:

[0012] Extract the text feature vectors of the interactive texts, determine the text feature similarity between the interactive texts based on the similarity between the text feature vectors of the interactive texts, and / or based on the time difference between the corresponding timestamps of the interactive texts in the target video, and cluster the interactive texts based on the text feature similarity.

[0013] Optionally, before clustering the interactive text based on the text feature similarity, the method further includes:

[0014] The interactive text is classified based on a pre-trained classification model, and interactive text in a preset category is removed; the preset categories include: specific behavior category and / or negative evaluation category.

[0015] Optionally, before clustering the interactive text based on the text feature similarity, the method further includes:

[0016] Remove non-ideographic characters from the interactive text.

[0017] Optionally, the step of determining the topic tags of the target video based on the clustering tags of the interactive text clusters includes:

[0018] The clustering label of each interactive text cluster is determined as the topic label of the target video;

[0019] or,

[0020] The number of interactive texts contained in each interactive text cluster is determined, and the clustering label of the interactive text cluster containing a number of interactive texts not less than a preset value is determined as the topic label of the target video.

[0021] In a second aspect of the present invention, a method for displaying a list of items is provided, applied to a client, comprising:

[0022] The system acquires video list search information and sends it to a server, so that the server determines a recommended video list that matches the video list search information. Each video in the recommended video list contains a topic tag with a text feature similarity to the search information that is not less than a preset similarity. The topic tag corresponding to each video is determined based on the clustering tag of the interactive text cluster obtained after clustering the interactive text published for that video.

[0023] Receive the recommended movie list returned by the server and display the recommended movie list.

[0024] In a third aspect of the invention, a video tag determination device is also provided, comprising:

[0025] The acquisition module is used to acquire interactive text posted in response to the target video.

[0026] The clustering module is used to cluster the interactive text based on the text feature similarity of the interactive text, to obtain one or more interactive text clusters and clustering labels for each interactive text cluster;

[0027] The determination module is used to determine the topic tags of the target video based on the clustering tags of the interactive text clusters.

[0028] In a fourth aspect of this invention, a device for displaying a list of items is provided, applied to a client, comprising:

[0029] An acquisition module is used to acquire video list search information and send the video list search information to a server, so that the server determines a recommended video list that matches the video list search information; wherein, each video in the recommended video list contains a topic tag with a text feature similarity of not less than a preset similarity with the search information, and the topic tag corresponding to each video is determined based on the clustering tag of the interactive text cluster obtained after clustering the interactive text published for that video;

[0030] The display module is used to receive the recommended video list returned by the server and display the recommended video list.

[0031] In a fifth aspect of the present invention, an electronic device is also provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus.

[0032] Memory, used to store computer programs;

[0033] The processor, when executing a program stored in memory, implements the video tag determination method or the video list information display method described above.

[0034] In another aspect of the present invention, a computer-readable storage medium is also provided, wherein a computer program is stored therein, and when the computer program is executed by a processor, it implements any of the video tag determination methods or video list information display methods described above.

[0035] In another aspect of the present invention, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the video tag determination methods or video list information display methods described above.

[0036] The video tag determination method, video list information display method, and related apparatus provided in this invention obtain interactive text posted for a target video, cluster the interactive text, and determine the topic tags of the target video based on the clustering tags of the interactive text clusters obtained from the clustering. It can track user-generated topics related to the video through interactive text, and obtain clustering tags that can be used to represent these related topics by clustering the interactive text. Therefore, the topic tags determined based on the clustering tags can effectively represent the topics that users are interested in regarding the target video. Compared to traditional video tags generated based on actors or themes, video topic tags are more intelligent, detailed, and diverse. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0038] Figure 1 This is a flowchart illustrating a video tag determination method provided in an embodiment of the present invention;

[0039] Figure 2 This is another flowchart illustrating the video tag determination method provided in this embodiment of the invention;

[0040] Figure 3 This is a schematic diagram of the video tag determination method provided in an embodiment of the present invention;

[0041] Figure 4 This is a flowchart illustrating the method for displaying item list information provided in an embodiment of the present invention;

[0042] Figure 5 This is a schematic diagram of the video tag determining device provided in an embodiment of the present invention;

[0043] Figure 6 This is a schematic diagram of the structure of the device for displaying item list information provided in an embodiment of the present invention;

[0044] Figure 7 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0045] The technical solutions of the present invention will now be described with reference to the accompanying drawings in the embodiments of the present invention.

[0046] To address the issue of insufficient diversity in conventional video tags, this invention provides a method for determining video tags. Figure 1 This is a flowchart illustrating a video tag determination method provided in an embodiment of the present invention. See also... Figure 1The video tag determination method provided in this embodiment of the invention specifically includes the following steps:

[0047] Step S101: Obtain the interactive text published for the target video.

[0048] Step S102: Based on the text feature similarity of the interactive texts, cluster the interactive texts to obtain one or more interactive text clusters and clustering labels for each interactive text cluster.

[0049] Step S103: Determine the topic tags of the target video based on the clustering tags of the interactive text clusters.

[0050] The video tag determination method provided in this invention obtains interactive text posted in response to a target video, clusters the interactive text, and determines the topic tags of the target video based on the clustering tags of the interactive text clusters obtained from the clustering. It can track user-generated topics related to the video through interactive text, and by clustering the interactive text, obtain clustering tags that can be used to represent these related topics. Therefore, the topic tags determined based on the clustering tags can effectively represent the topics that users are interested in regarding the target video. Compared to traditional video tags generated based on actors or themes, video topic tags are more intelligent, detailed, and diverse.

[0051] The steps described above in the embodiments of the present invention will be explained separately below:

[0052] In step S101, the target video is the video that needs to be tagged with a video tag.

[0053] Specifically, during or after watching a video, users typically post various text messages to share their evaluations or emotional experiences, and to interact with other users. These text messages are the interactive text in this embodiment of the invention. For example, user-generated comments and bullet screen messages on a video constitute interactive text.

[0054] In this embodiment of the invention, interactive text published for a target video can be obtained through platforms such as video websites and video review websites.

[0055] In step S102, the interactive texts are clustered. Specifically, the similarity between any two interactive texts in terms of text features can be calculated. Interactive texts with high similarity between their text features are considered as an interactive text cluster, thereby obtaining one or more interactive text clusters.

[0056] As an example, in the process of clustering interactive text based on the text feature similarity of interactive text, the interactive text can first be converted into text feature vectors, and then the interactive text can be clustered based on the similarity between the text feature vectors of the interactive text.

[0057] In this embodiment of the invention, clustering algorithms can be used to cluster interactive text. The specific clustering algorithm can be selected based on actual needs. As an example, k-means (a clustering algorithm) or kNN (a clustering algorithm) can be used to cluster interactive text.

[0058] Specifically, after clustering, the interactive texts in each interactive text cluster have a high degree of text feature similarity. Therefore, interactive texts within the same interactive text cluster tend to express similar semantics, and thus, interactive texts within the same interactive text cluster can be considered as discussions on the same topic.

[0059] Based on this, for each interactive text cluster determined in step S102, a clustering label can be determined to represent the topic discussed in the interactive texts within that cluster. In this embodiment of the invention, this label is specifically text; for example, the label can be a word or a phrase. As an example, the label may specifically represent characters or plot points in a video.

[0060] The specific method for determining the clustering labels of interactive text clusters can be found in related technologies, and this embodiment of the invention does not limit this method. As an example, labels for interactive text clusters can be determined using keyword generation algorithms or keyword extraction algorithms.

[0061] Specifically, for each interactive text cluster identified in step S102, a clustering label for that interactive text cluster can be determined.

[0062] In step S103, clustering tags of one or more interactive text clusters can be selected as topic tags for the target video according to actual needs. This embodiment of the invention does not limit the specific method and scope of selecting interactive text clusters.

[0063] As an example, the cluster labels of all interactive text clusters obtained by clustering can be used as the topic labels of the target video. As another example, the cluster labels of only one interactive text cluster can be used as the topic labels of the target video.

[0064] The video tag determination method provided in this invention obtains interactive text posted in response to a target video, clusters the interactive text, and determines the topic tags of the target video based on the clustering tags of the interactive text clusters obtained from the clustering. It can track user-generated topics related to the video through interactive text, and by clustering the interactive text, obtain clustering tags that can be used to represent these related topics. Therefore, the topic tags determined based on the clustering tags can effectively represent the topics that users are interested in regarding the target video. Compared to traditional video tags generated based on actors or themes, video topic tags are more intelligent, detailed, and diverse.

[0065] In one embodiment of the present invention, the aforementioned step of clustering interactive text based on text feature similarity includes:

[0066] Extract text feature vectors from interactive texts, determine text feature similarity between interactive texts based on the similarity between text feature vectors and / or the time difference between corresponding timestamps of interactive texts in the target video, and cluster interactive texts based on text feature similarity.

[0067] Specifically, text feature vectors can represent the text features of interactive text in a vectorized form. Therefore, the similarity between interactive texts in terms of text feature vectors can be used to accurately represent the text feature similarity between interactive texts. The methods for determining the text feature vectors of interactive texts and calculating the similarity between text feature vectors can refer to relevant technologies. This embodiment of the invention does not limit these methods. As an example, text feature vectors of interactive texts can be extracted using TF-IDF (Term Frequency / Inverse Document Frequency, a text feature extraction algorithm) or word2vec (a text feature extraction algorithm), and the similarity between text feature vectors can be determined based on the cosine distance or Euclidean distance between them.

[0068] Furthermore, in practical applications, some interactive text has a corresponding timestamp in the video. Taking bullet comments as an example, bullet comments are displayed when the video plays to a specific image frame. Therefore, the time point in the video where the bullet comment is displayed can be understood as the corresponding timestamp of the bullet comment in the target video.

[0069] In practical applications, users post interactive text based on the actual content played in the video. Therefore, comments with similar timestamps in the target video—that is, comments corresponding to the same or similar image frames in the target video—tend to discuss the events corresponding to those image frames, and thus usually represent the same topic. Therefore, if the time difference between the corresponding timestamps of the interactive text in the target video is less than a threshold, meaning the corresponding timestamps in the target video are relatively close, these interactive texts can be considered to have a high degree of similarity in textual features.

[0070] Therefore, in this embodiment of the invention, if the time difference between the corresponding timestamps of the interactive texts in the target video is small, these interactive texts can also be grouped into the same interactive text cluster.

[0071] In this embodiment of the invention, the text feature vector of the interactive text and / or the corresponding timestamp of the interactive text in the target video can be used as the basis for determining the text feature similarity of the interactive text, according to actual needs. This invention does not limit this. As an example, the text feature similarity between interactive texts can be determined solely based on the text feature vector of the interactive text or the corresponding timestamp of the interactive text in the target video. Alternatively, different weights can be assigned to the text feature vector of the interactive text and its corresponding timestamp in the video, and then the two can be combined to determine the text feature similarity of the interactive text.

[0072] In this embodiment of the invention, the text feature similarity between interactive texts is determined based on the similarity between the text feature vectors of the interactive texts and / or the time difference between the corresponding timestamps of the interactive texts in the target video. Compared with traditional text similarity algorithms that only consider text feature vectors, this method uses the time difference between the corresponding timestamps of the interactive texts in the target video as the basis for determining text feature similarity, based on the actual characteristics of the interactive texts. This results in greater comprehensiveness in determining the text feature similarity between interactive texts, which helps to improve the clustering effect of interactive texts and thus improve the accuracy of the determined topic tags of the target video.

[0073] In one embodiment of the present invention, the step of determining the topic tag of the target video based on the clustering tag of interactive text clusters includes:

[0074] The clustering labels of each interactive text cluster are determined as the topic labels of the target video;

[0075] or,

[0076] Determine the number of interactive texts contained in each interactive text cluster, and determine the clustering labels of interactive text clusters containing at least a preset number of interactive texts as the topic labels of the target video.

[0077] In this embodiment of the invention, the clustering label of each interactive text cluster obtained by clustering can be used as the topic label of the target video. In this case, each topic label can more comprehensively represent the topic related to the target video.

[0078] In this embodiment of the invention, the number of interactive texts contained in each interactive text cluster can be determined first, and only the clustering labels of interactive text clusters containing more interactive texts than a preset value can be determined as the topic labels of the target video.

[0079] Specifically, if an interactive text cluster contains a small number of interactive texts, the meaning expressed by these texts may deviate from the actual topic of the target video. Therefore, the clustering label of this interactive text cluster cannot be used to represent the topic related to the target video. Conversely, if an interactive text cluster contains a large number of interactive texts, it indicates that many users will discuss the target video based on the topic corresponding to this interactive text cluster. Therefore, the clustering label of this interactive text cluster can be used to represent the topic of the target video.

[0080] Therefore, only the cluster labels of interactive text clusters containing a large amount of interactive text can be used as the topic labels of the target video. In this case, the selected cluster labels of interactive text can be considered as topics with a certain degree of discussion about the target video, thus making these cluster labels more representative as topic labels of the target video.

[0081] It should be understood that, based on the number of interactive texts contained in each interactive text cluster, the interactive text clusters can be sorted in descending order, and the clustering tags of a preset proportion or a preset number of interactive text clusters can be used as the topic tags of the target video. This approach is essentially the same as the example described above in the embodiments of the present invention.

[0082] As mentioned earlier, video tags can be applied to video recommendation tasks. Current video recommendation methods mainly rely on actors and genres, making it difficult to leverage personalized information shared by users. Furthermore, many shows suddenly generate new memes long after they've finished airing. Conventional video recommendation methods cannot track these video-related memes on a large scale, nor can they recommend videos based on these user-generated memes, resulting in a relatively limited selection of content to recommend to users.

[0083] To address this issue, after determining the video's topic tags based on the above embodiments of the present invention, a video list can be generated based on these video topics, thereby enabling more personalized and diverse video recommendations to users. Figure 2This is another flowchart illustrating the video tag determination method provided in this embodiment of the invention. See [link / reference]. Figure 2 In one embodiment of the present invention, the video tag determination method provided by the present invention further includes the following steps:

[0084] Step S104: Determine the text similarity between topic tags identified for multiple target videos, aggregate the target videos corresponding to topic tags with text similarity not less than the similarity threshold, and obtain a video recommendation list.

[0085] Through the aforementioned steps S101-S103, one or more video tags can be determined for each target video.

[0086] In this step, topic tags identified for multiple target videos can be obtained, the text similarity of these topic tags can be calculated, and then the target videos corresponding to topic tags with text similarity not less than the similarity threshold can be included in the same video recommendation list, or video list.

[0087] In this embodiment of the invention, the text similarity between topic tags can be calculated using a text similarity algorithm. Specifically, topic tags can first be converted into text feature vectors, and then the similarity between the text feature vectors can be calculated to obtain the text similarity between topic tags. The text similarity algorithm used in this process can be selected based on actual needs, and this embodiment of the invention does not limit it.

[0088] As an example, the SimHash algorithm (a text similarity algorithm) or the BERT model can be used to calculate the text similarity between topic tags.

[0089] The similarity threshold can be configured according to actual needs. For example, the similarity threshold can be set to 1, in which case target videos with the same topic tags will be included in a video recommendation list. Alternatively, the similarity threshold can be set to a value less than 1, in which case target videos with similar topic tags will be included in a video recommendation list.

[0090] As an example, if topic tags A and B are identified for video 1, topic tags A and C for video 2, and topic tags D and E for video 3, then with a similarity threshold of 1, both video 1 and video 2 have topic tag A and are therefore included in a video recommendation list. If the similarity threshold is set to a value less than 1, then if the similarity between topic tags C and D is not less than the similarity threshold, then video 2 and video 3 can be included in a video recommendation list.

[0091] Once videos with the same or similar topic tags are included in a video recommendation list, the videos in the recommendation list can be grouped together on video websites for users to watch, or the video recommendation list can be pushed to users, so that users interested in the relevant topics can more easily find and watch videos of interest through the video recommendation list.

[0092] In this embodiment of the invention, based on the more diverse and personalized topic tags, by including videos with the same or similar topic tags in a video recommendation list, when applied to video recommendations on video websites, personalized recommendations can be made for users with different interests, thereby increasing user engagement time and improving user retention.

[0093] In one embodiment of the present invention, the obtained interactive text may be filtered first, and then only the filtered interactive text may be clustered. In this embodiment of the present invention, before the aforementioned step S102, the method further includes:

[0094] The interactive text is classified based on a pre-trained classification model, and interactive text in a preset category is removed; the preset categories include: specific behavior category and / or negative evaluation category.

[0095] In practical applications, some interactive texts published in response to a target video are not actually related to the target video. If clustering is performed based on these interactive texts, the clustering labels of the resulting interactive text clusters cannot reflect the topics related to the target video. The topic labels determined on this basis are also not of reference value and the accuracy of the determined topic labels is insufficient.

[0096] To avoid this part of the interactive text interfering with the process of determining the topic tags of the target video, in this embodiment of the invention, interactive text of a specific behavior category can be determined according to actual needs and cannot be used to represent the topic of the target video, and this type of interactive text can be removed.

[0097] It should be understood that a specific behavior category may include multiple subcategories, and the specific subcategories can be determined based on the actual situation. As an example, the interactive text for a specific behavior category may include: interactive text at the beginning and end, such as "XXX, I'm here" or "XXX, I'm leaving"; interactive text posted on a specific date, such as "Come see me on [Month] [Day]"; and meaningless interactive text, i.e., interactive text composed of randomly spelled letters or gibberish. It should be understood that the specific behavior categories listed here are only examples, and the interactive text for specific behavior categories that can be excluded in actual applications is not limited to these.

[0098] Furthermore, embodiments of the present invention can also remove interactive text in the negative evaluation category. Interactive text in the negative evaluation category refers to interactive text that makes negative evaluations of the video. For example, interactive text containing abusive language or sensitive words may be interactive text in the negative evaluation category.

[0099] Specifically, since video tags are typically used for video retrieval and recommendation, if interactive texts are clustered based on negative review categories, the resulting tags for the interactive text clusters cannot objectively reflect the video content, nor can they guide users interested in the topic to find the video.

[0100] Therefore, interactive texts of specific behavior categories and negative evaluation categories are not effective in determining topic tags. Thus, these types of interactive texts can be removed as noisy interactive texts. The remaining categories of interactive texts are normal interactive texts, which can be clustered to determine the topic tags of the target video.

[0101] The specific subcategories that the specific behavior categories to be removed can be pre-determined, and an interactive text corpus for training can be obtained. The categories of the interactive texts in the corpus are then labeled to obtain interactive texts labeled as one or more specific behavior categories, as well as interactive texts labeled as negative evaluation categories. Based on these labeled interactive texts, a classification model is trained to obtain a classifier used in this embodiment of the invention to identify interactive texts of specific behavior categories and negative evaluation categories. The specific classification algorithm and training method used by the classifier can be selected based on actual needs, and this embodiment of the invention does not limit this.

[0102] As an example, BERT (Bidirectional Encoder Representations from Transformers, a language model) or GPT (Generative Pre-Training, a language model) can be used as classifiers to classify interactive text.

[0103] In this embodiment of the invention, by classifying interactive text using a classification model and removing interactive text of specific behavioral categories and / or negative evaluation categories, it is possible to ensure that the interactive text used to determine the topic tags of the video can effectively represent the topics related to the video content, thereby improving the effectiveness of the determined topic tags.

[0104] In one embodiment of the present invention, the acquired interactive text may be preprocessed, and then the processed interactive text may be clustered. In this embodiment of the present invention, before the aforementioned step S102, the method further includes:

[0105] Remove non-ideographic characters from interactive text.

[0106] In this embodiment of the invention, each interactive text may be composed of characters, some of which are not relevant to the personalized content the user intends to express and can be considered as non-ideographic characters. The specific categories of characters included in non-ideographic characters can be selected according to actual needs, and this embodiment of the invention does not limit this. As an example, non-ideographic characters may include: topics from the video platform used by the interactive text, such as "#XXX10thAnniversary#"; emoticons; foreign languages; and invalid symbols, such as "@#￥%".

[0107] As an example, if the interactive text before preprocessing is "#XXX10thAnniversary#XXXIsSoHandsome[Smile]", then the interactive text obtained after removing non-ideographic characters is "XXXIsSoHandsome".

[0108] In practical applications, specific rules for removing non-ideal characters can be determined based on actual needs. Filters can be written based on these rules, and interactive text can be filtered using these filters to obtain the interactive text after removing non-ideal characters. As an example, filters can be written with specific rules for removing words of a particular part of speech, specific symbols, or strings with a specific format.

[0109] In this embodiment of the invention, by preprocessing the interactive text and removing non-ideographic characters, interference from these non-ideographic characters can be avoided when clustering the interactive text and determining the topic tags of the video based on the interactive text, thereby improving the effectiveness of the determined topic tags and improving the computational efficiency in the process of determining topic tags.

[0110] Figure 3 This is a schematic diagram of the video tag determination method provided in the embodiments of the present invention. The following is in conjunction with... Figure 3 The video tag determination method provided in the embodiments of the present invention will be described by way of example.

[0111] First, obtain the original interactive text posted by the user. Specifically, the original interactive text can be the user's original bullet comments / comments. See examples of some original interactive texts. Figure 3 Part (a) of the text.

[0112] After obtaining the original interactive text posted by users, text preprocessing is required to remove non-ideal characters. Specifically, invalid symbols, foreign languages, topics, and emoticons can be removed from each piece of original interactive text. See also Figure 3 Part (b) shows the preprocessed interactive text, which reveals that invalid symbols, foreign languages, emoticons, and other information in the original interactive text have been removed.

[0113] After preprocessing the interactive text, it needs to be categorized, retaining only the normal text. Taking bullet comments as an example, this involves removing noisy bullet comments without information and keeping the normal ones. The specific categories of noisy bullet comments can be found in the aforementioned examples. See also... Figure 3 In part (c), the first bullet comment is identified as a normal bullet comment, and the second bullet comment is identified as a noise bullet comment, so the second bullet comment is removed.

[0114] After preprocessing the interactive text and preserving the normal text, the preserved interactive text can be clustered and integrated. For each clustered interactive text cluster, its clustering topic, i.e., its tag, can be determined. Taking bullet comments as an example, see [link to example]. Figure 3 In part (d), the left side shows the clustering topics of the interactive text clusters, and the right side shows the specific bullet comments contained in the interactive text clusters. It can be seen that the clustering topics can effectively represent the topics that the interactive texts within the interactive text clusters focus on.

[0115] After clustering the interactive texts, you can choose to define the clustering topics of the interactive text clusters as video topic tags, and then add videos with the same or similar topic tags to a video recommendation list, thereby enabling personalized video recommendations for users with different interests.

[0116] Based on the same inventive concept, this invention also provides a method for displaying a list of items, specifically applied to a client-side application. Figure 4 This is a flowchart illustrating the method for displaying item list information provided in an embodiment of the present invention. See also... Figure 4 The method specifically includes the following steps:

[0117] Step S401: Obtain video list search information and send the video list search information to the server so that the server can determine recommended video lists that match the video list search information; wherein, each video in the recommended video list contains a topic tag with a text feature similarity of not less than a preset similarity with the search information, and the topic tag corresponding to each video is determined based on the clustering tag of the interactive text cluster obtained after clustering the interactive text published for that video.

[0118] Step S402: Receive the recommended movie list returned by the server and display the recommended movie list.

[0119] In this embodiment of the invention, the video list search information specifically refers to the text entered by the user in the client. This text can be used to search for video lists or videos. Specifically, for this video list search information, the text feature similarity between the video list search information and the topic tags of multiple videos can be calculated. For videos containing topic tags with a text feature similarity of not less than a preset similarity with the video list search information, these videos can be included in the recommended video list based on the video list search information. In this embodiment of the invention, the preset similarity can be set based on actual needs.

[0120] The topic tags for each video can be found in the descriptions of any of the foregoing embodiments of this application, and will not be repeated here.

[0121] Based on the fact that video hashtags can represent personalized information related to the video in a diverse and comprehensive way, the text similarity between video list search information and hashtags can be used to determine recommended video lists that match the video list search information. This also allows for a more comprehensive and diverse display of videos related to the video list search information, making it easier for users to find videos of interest based on the video list search information. This helps to increase user engagement time on the client and improve user retention.

[0122] Based on the same inventive concept, embodiments of the present invention also provide a video tag determining device, see [link to relevant documentation]. Figure 5 The device includes:

[0123] Module 501 is used to acquire interactive text published for the target video.

[0124] Clustering module 502 is used to cluster interactive texts based on the text feature similarity of interactive texts, and obtain one or more interactive text clusters and clustering labels for each interactive text cluster;

[0125] Module 503 determines the topic tags of the target video based on the clustering tags of the interactive text clusters.

[0126] The video tag determination device provided in this invention obtains interactive text posted for a target video, clusters the interactive text, and determines the topic tags of the target video based on the clustering tags of the interactive text clusters obtained from the clustering. It can track user-generated topics related to the video through interactive text, and obtain clustering tags that can represent these related topics by clustering the interactive text. Therefore, the topic tags determined based on the clustering tags can effectively represent the topics that users are interested in regarding the target video. Compared to traditional video tags generated based on actors or themes, video topic tags are more intelligent, detailed, and diverse.

[0127] In one embodiment of the present invention, the video tag determination device further includes:

[0128] The aggregation module is used to determine the text similarity between topic tags identified for multiple target videos, and aggregates the target videos corresponding to topic tags with text similarity not less than the similarity threshold to obtain a video recommendation list.

[0129] In one embodiment of the present invention, the clustering module 502 is specifically used for:

[0130] Extract text feature vectors from interactive texts, determine text feature similarity between interactive texts based on the similarity between text feature vectors and / or the time difference between corresponding timestamps of interactive texts in the target video, and cluster interactive texts based on text feature similarity.

[0131] In one embodiment of the present invention, the video tag determination device further includes:

[0132] The first elimination module is used to classify interactive text based on a pre-trained classification model and eliminate interactive text in preset categories; the preset categories include: specific behavior categories and / or negative evaluation categories.

[0133] In one embodiment of the present invention, the video tag determination device further includes:

[0134] The second removal module is used to remove non-ideographic characters from interactive text.

[0135] In one embodiment of the present invention, the determining module 503 is specifically used for:

[0136] The clustering labels of each interactive text cluster are determined as the topic labels of the target video;

[0137] or,

[0138] Determine the number of interactive texts contained in each interactive text cluster, and determine the clustering labels of interactive text clusters containing at least a preset number of interactive texts as the topic labels of the target video.

[0139] Based on the same inventive concept, this invention also provides a method for displaying a list of items, applied to a client-side application. See [link to relevant documentation]. Figure 6 The device includes:

[0140] The acquisition module 601 is used to acquire video list search information and send the video list search information to the server so that the server can determine a recommended video list that matches the video list search information; wherein, each video in the recommended video list contains a topic tag with a text feature similarity of not less than a preset similarity with the search information, and the topic tag corresponding to each video is determined based on the clustering tag of the interactive text cluster obtained after clustering the interactive text published for that video;

[0141] Display module 602 is used to receive the recommended movie list returned by the server and display the recommended movie list.

[0142] Based on the fact that video hashtags can represent personalized information related to the video in a diverse and comprehensive way, the text similarity between video list search information and hashtags can be used to determine recommended video lists that match the video list search information. This also allows for a more comprehensive and diverse display of videos related to the video list search information, making it easier for users to find videos of interest based on the video list search information. This helps to increase user engagement time on the client and improve user retention.

[0143] This invention also provides an electronic device, such as... Figure 7 As shown, it includes a processor 701, a communication interface 702, a memory 703, and a communication bus 704, wherein the processor 701, the communication interface 702, and the memory 703 communicate with each other through the communication bus 704.

[0144] Memory 703 is used to store computer programs;

[0145] When processor 701 executes a program stored in memory 703, it performs the following steps:

[0146] Retrieve interactive text posted for the target video;

[0147] Based on the text feature similarity of interactive texts, the interactive texts are clustered to obtain one or more interactive text clusters and cluster labels for each interactive text cluster.

[0148] Based on the clustering tags of interactive text clusters, the topic tags of the target video are determined.

[0149] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This communication bus can be divided into address bus, data bus, control bus, etc. For ease of illustration, only one thick line is used to represent it in the diagram, but this does not mean that there is only one bus or one type of bus.

[0150] The communication interface is used for communication between the aforementioned terminal and other devices.

[0151] The memory may include random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor.

[0152] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0153] In another embodiment of the present invention, a computer-readable storage medium is also provided, wherein a computer program is stored therein, and when the computer program is executed by a processor, it implements the video tag determination method described in any of the above embodiments.

[0154] In another embodiment of the present invention, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute any of the video tag determination methods described in the above embodiments.

[0155] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid state disk (SSD)).

[0156] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0157] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on its differences from other embodiments. In particular, the embodiments for the video tag determining device, electronic device, and readable storage medium are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0158] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of protection of the present invention.

Claims

1. A method of video tag determination, the method comprising: include: Retrieve interactive text posted for the target video; Based on the text feature similarity of the interactive texts, the interactive texts are clustered to obtain one or more interactive text clusters and clustering labels for each interactive text cluster; Based on the clustering tags of the interactive text clusters, the topic tags of the target video are determined; The step of clustering the interactive text based on the text feature similarity of the interactive text includes: Based on the time difference between the corresponding timestamps of the interactive text in the target video, the text feature similarity between the interactive texts is determined, and the interactive texts are clustered based on the text feature similarity. The smaller the time difference between the corresponding timestamps of the interactive text in the target video, the higher the text feature similarity of the interactive texts. Interactive texts with a time difference between the corresponding timestamps of the interactive text in the target video that is less than a threshold are grouped into the same interactive text cluster. The step of determining the topic tag of the target video based on the clustering tag of the interactive text cluster includes: determining the number of interactive texts contained in each interactive text cluster, and determining the clustering tag of the interactive text cluster containing a number of interactive texts not less than a preset value as the topic tag of the target video.

2. The method of claim 1, wherein, The method further includes: The text similarity between the topic tags determined for multiple target videos is determined, and the target videos corresponding to the topic tags whose text similarity is not less than the similarity threshold are aggregated to obtain a video recommendation list.

3. The method according to any one of claims 1-2, characterized in that, Before clustering the interactive text based on its text feature similarity, the method further includes: The interactive text is classified based on a pre-trained classification model, and interactive text in a preset category is removed; the preset categories include: specific behavior category and / or negative evaluation category.

4. The method according to any one of claims 1-2, characterized in that, Before clustering the interactive text based on its text feature similarity, the method further includes: Remove non-ideographic characters from the interactive text.

5. A method of presenting a piece of information, characterized by Applied to a client, the method includes: The system acquires video list search information and sends it to a server, so that the server determines a recommended video list that matches the video list search information. Each video in the recommended video list contains a topic tag with a text feature similarity to the search information that is not less than a preset similarity. The topic tag corresponding to each video is determined based on the clustering tag of the interactive text cluster obtained after clustering the interactive text published for that video. Receive the recommended video list returned by the server and display the recommended video list; The interactive text posted in response to the video was clustered, including: Based on the time difference between the corresponding timestamps of the interactive text in the video, the text feature similarity between the interactive texts is determined, and the interactive texts are clustered based on the text feature similarity. The smaller the time difference between the corresponding timestamps of the interactive text in the video, the higher the text feature similarity of the interactive texts. Interactive texts with a time difference between the corresponding timestamps in the video that is less than a threshold are grouped into the same interactive text cluster. The topic tags corresponding to the video were determined in the following way: The number of interactive texts contained in each interactive text cluster is determined, and the clustering label of the interactive text cluster containing a number of interactive texts not less than a preset value is determined as the topic label of the video.

6. A video tag determining device, characterized in that, include: The acquisition module is used to acquire interactive text posted in response to the target video. The clustering module is used to cluster the interactive text based on the text feature similarity of the interactive text, to obtain one or more interactive text clusters and clustering labels for each interactive text cluster; The determination module is used to determine the topic tags of the target video based on the clustering tags of the interactive text clusters; The clustering module is specifically used to: determine the text feature similarity between the interactive texts based on the time difference between the corresponding timestamps of the interactive texts in the target video, and cluster the interactive texts based on the text feature similarity, wherein the smaller the time difference between the corresponding timestamps of the interactive texts in the target video, the higher the text feature similarity of the interactive texts, and to group interactive texts with a time difference between the corresponding timestamps in the target video that is less than a threshold into the same interactive text cluster; The determining module is specifically used to: determine the number of interactive texts contained in each interactive text cluster, and determine the clustering label of the interactive text cluster containing a number of interactive texts not less than a preset value as the topic label of the target video.

7. A device for displaying a list of items, characterized in that, Applied to the client side, including: An acquisition module is used to acquire video list search information and send the video list search information to a server, so that the server determines a recommended video list that matches the video list search information; wherein, each video in the recommended video list contains a topic tag with a text feature similarity of not less than a preset similarity with the search information, and the topic tag corresponding to each video is determined based on the clustering tag of the interactive text cluster obtained after clustering the interactive text published for that video; The display module is used to receive the recommended video list returned by the server and display the recommended video list; The interactive text posted in response to the video was clustered, including: Based on the time difference between the corresponding timestamps of the interactive text in the video, the text feature similarity between the interactive texts is determined, and the interactive texts are clustered based on the text feature similarity. The smaller the time difference between the corresponding timestamps of the interactive text in the video, the higher the text feature similarity of the interactive texts. Interactive texts with a time difference between the corresponding timestamps in the video that is less than a threshold are grouped into the same interactive text cluster. The topic tags corresponding to the video were determined in the following way: The number of interactive texts contained in each interactive text cluster is determined, and the clustering label of the interactive text cluster containing a number of interactive texts not less than a preset value is determined as the topic label of the video.

8. An electronic device, characterized in that, It includes a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; Memory, used to store computer programs; A processor, when executing a program stored in memory, implements the steps of the method described in any one of claims 1-5.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the method described in any one of claims 1-5.

Images