Media data determination method, device, electronic device and storage medium

Abstract

The present disclosure relates to a method, device, electronic device, and storage medium for determining media data. The method includes: obtaining a first media data set and a second media data set, wherein the first media data set includes a plurality of first media data received within a preset time, and the second media data set includes a plurality of second media data, wherein the second media data carries user behavior data; determining first target media data in the first media data set that matches third media data in a media database; obtaining second target media data in the second media data set that has target dynamic indicator characteristics based on the user behavior data; and determining the first target media data, the second target media data, and the third media data as a target data set. The present disclosure ensures the richness of the determined media data and solves the problem of a single data source for determining media data.

Description

Technical Field

[0001] The present disclosure relates to the field of Internet technology, and in particular to a method and device for determining media data, an electronic device, and a storage medium. Background Art

[0002] With the rapid development of media applications such as the Internet and video in recent years, the consumption demand for media data is changing with each passing day. The site can better carry out personalized and customized media data acquisition and recommendation to meet user needs.

[0003] However, in specific demand scenarios, such as promoting media applications, it is impossible to make personalized recommendations for a certain user. Instead, similar media data can only be determined based on existing media data, and the data source for determining the media data is relatively single. Summary of the Invention

[0004] The present disclosure provides a method, device, electronic device, and storage medium for determining media data to at least solve the problem of a single data source for determining media data in related technologies. The technical solutions of the present disclosure are as follows:

[0005] According to a first aspect of an embodiment of the present disclosure, a method for determining media data is provided, including:

[0006] Obtaining a first media data set and a second media data set, where the first media data set includes a plurality of first media data received within a preset time, and the second media data set includes a plurality of second media data, where the second media data carries user behavior data;

[0007] Determining first target media data in the first media data set that matches third media data in a media database, where the third media data is media data that meets a preset condition;

[0008] obtaining, based on the user behavior data, second target media data having a target dynamic indicator feature from the second media data set, the target dynamic indicator feature being used to characterize a dynamic indicator feature corresponding to the obtained second target media data, the dynamic indicator feature being a feature associated with the user behavior data;

[0009] The first target media data, the second target media data, and the third media data are determined as a target data set.

[0010] Optionally, determining first target media data in the first media data set that matches third media data in the media database includes:

[0011] performing feature extraction on the first media data in the first media data set to obtain first media data features;

[0012] Searching a search library according to the first media data feature, wherein the search library includes the third media data feature corresponding to the third media data;

[0013] If a third media data feature matching the first media data feature exists in the search library, the first media data is determined to be the first target media data.

[0014] Optionally, searching a search library according to the first media data feature includes:

[0015] determining a first similarity between the first media data feature and each third media data feature in the search library;

[0016] If there is a third media data feature whose first similarity is greater than or equal to a first similarity threshold, it is determined that there is a third media data feature matching the first media data feature in the search library.

[0017] Optionally, before searching in a search library according to the first media data feature, the method further includes:

[0018] performing feature extraction on each of the third media data in the media database to obtain third media data features corresponding to each of the third media data;

[0019] The search library is constructed according to the third media data feature corresponding to each third media data.

[0020] Optionally, obtaining, based on the user behavior data, second target media data having target dynamic indicator characteristics in the second media data set includes:

[0021] Obtaining user behavior data of the second media data in the second media data set, and obtaining a target dynamic indicator feature vector corresponding to the target dynamic indicator feature;

[0022] determining a dynamic indicator feature vector of the second media data based on user behavior data of the second media data;

[0023] Determining a second similarity between the dynamic indicator feature vector and the target dynamic indicator feature vector;

[0024] If the second similarity is greater than or equal to a second similarity threshold, the second media data is determined to be the second target media data.

[0025] Optionally, determining the first target media data, the second target media data, and the third media data as a target data set includes:

[0026] Sending the first target media data and the second target media data to a manual review system, and obtaining a manual review result;

[0027] adding the first target media data and the second target media data whose manual review results are approved to the media database;

[0028] The first target media data and the second target media data that have passed the review in the media database, as well as the third media data, are determined as the target data set.

[0029] Optionally, the target data set is displayed on an interface of a target application.

[0030] According to a second aspect of an embodiment of the present disclosure, a media data determination apparatus is provided, including:

[0031] a data set acquisition module configured to acquire a first media data set and a second media data set, wherein the first media data set includes a plurality of first media data received within a preset time period, and the second media data set includes a plurality of second media data, wherein the second media data carries user behavior data;

[0032] a first target data determining module configured to determine first target media data in the first media data set that matches third media data in the media database, where the third media data is media data that meets a preset condition;

[0033] a second target data acquisition module configured to acquire, from the second media data set, second target media data having a target dynamic indicator feature based on the user behavior data, wherein the target dynamic indicator feature is used to represent a dynamic indicator feature corresponding to the acquired second target media data, the dynamic indicator feature being a feature associated with the user behavior data;

[0034] The target data set determining module is configured to determine the first target media data, the second target media data, and the third media data as a target data set.

[0035] Optionally, the first target data determination module includes:

[0036] a feature extraction unit, configured to perform feature extraction on the first media data in the first media data set to obtain first media data features;

[0037] a retrieval unit configured to perform a search in a retrieval library based on the first media data feature, wherein the retrieval library includes a third media data feature corresponding to the third media data;

[0038] The first target data determining unit is configured to determine that the first media data is the first target media data if a third media data feature matching the first media data feature exists in the search library.

[0039] Optionally, the retrieval unit is configured to execute:

[0040] determining a first similarity between the first media data feature and each third media data feature in the search library;

[0041] If there is a third media data feature whose first similarity is greater than or equal to a first similarity threshold, it is determined that there is a third media data feature matching the first media data feature in the search library.

[0042] Optionally, the device further includes:

[0043] a seed data feature extraction module configured to perform feature extraction on each of the third media data in the media database to obtain a third media data feature corresponding to each of the third media data;

[0044] The retrieval library construction module is configured to construct the retrieval library according to the third media data feature corresponding to each third media data.

[0045] Optionally, the second target data acquisition module includes:

[0046] A user behavior data acquisition unit configured to acquire user behavior data of the second media data in the second media data set, and acquire a target dynamic indicator feature vector corresponding to the target dynamic indicator feature;

[0047] a dynamic feature vector determining unit configured to determine a dynamic indicator feature vector of the second media data based on user behavior data of the second media data;

[0048] a similarity determination unit configured to determine a second similarity between the dynamic indicator feature vector and the target dynamic indicator feature vector;

[0049] The second target data determining unit is configured to determine that the second media data is the second target media data if the second similarity is greater than or equal to a second similarity threshold.

[0050] Optionally, the recalled data set determination module includes:

[0051] a manual review unit, configured to send the first target media data and the second target media data to a manual review system and obtain a manual review result;

[0052] A seed database data adding unit is configured to add the first target media data and the second target media data, whose manual review results are approved, to the media database;

[0053] The target data set determining unit is configured to determine the first target media data and the second target media data that have passed the review in the media database, as well as the third media data, as the target data set.

[0054] Optionally, the device further includes:

[0055] The data display module is used to display the target data set in the interface of the target application.

[0056] According to a third aspect of an embodiment of the present disclosure, there is provided an electronic device, including:

[0057] processor;

[0058] a memory for storing instructions executable by the processor;

[0059] The processor is configured to execute the instructions to implement the media data determination method as described in the first aspect.

[0060] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided. When instructions in the computer storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the media data determination method as described in the first aspect.

[0061] According to a fifth aspect of an embodiment of the present disclosure, a computer program product is provided, comprising a computer program or computer instructions, which, when executed by a processor, implements the media data determination method described in the first aspect.

[0062] The technical solutions provided by the embodiments of the present disclosure bring at least the following beneficial effects:

[0063] The present disclosure obtains a first media data set and a second media data set, determines first target media data in the first media data set that matches third media data in a media database, obtains second target media data in the second media data set that has target dynamic indicator characteristics based on user behavior data of the second media data, and determines the first target media data, the second target media data, and the third media data in the media database as the target data set. Since the characteristics of the media data itself are used to determine the first target media data, and the dynamic characteristics of the media data are used to determine the second target media data, the richness of the determined media data is guaranteed, and the problem of a single data source for determining the media data is solved.

[0064] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] The accompanying drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute an improper limitation of the present disclosure.

[0066] Figure 1 is a flow chart showing a method for determining media data according to an exemplary embodiment;

[0067] Figure 2 is a flow chart showing a method for determining media data according to an exemplary embodiment;

[0068] Figure 3 is a schematic diagram of an implementation method of the media data determination method in an embodiment of the present disclosure;

[0069] Figure 4 is a block diagram showing a device for determining media data according to an exemplary embodiment;

[0070] Figure 5 It is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0071] In order to enable ordinary people in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.

[0072] It should be noted that the terms "first," "second," and the like in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the numbers used in this manner are interchangeable where appropriate so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure as detailed in the appended claims.

[0073] Figure 1 FIG. 1 is a flow chart showing a method for determining media data according to an exemplary embodiment. Figure 1 As shown, the media data determination method is used in electronic devices such as servers, and includes the following steps.

[0074] In step S11, a first media data set and a second media data set are obtained, wherein the first media data set includes a plurality of first media data received within a preset time, and the second media data set includes a plurality of second media data, and the second media data carries user behavior data.

[0075] The first media data set may include one or more first media data. Each first media data set is media data received within a preset time period, is undisplayed media data, and does not have user behavior data. The first media data set is also newly uploaded media data and does not yet have user behavior data (e.g., browsing behavior, forwarding behavior, etc.). The second media data set may include one or more second media data sets. Each second media data set carries user behavior data, meaning that each second media data set has been viewed by a user. The first media data and the second media data may be videos, images, audio, etc.

[0076] The media data determination method can be triggered at a fixed time or according to an instruction. When the fixed time is reached or a data determination instruction is received, a first media data set consisting of first media data without user behavior data and a second media data set consisting of second media data carrying user behavior data are obtained.

[0077] In step S12, first target media data in the first media data set that matches the media data in the media database is determined, and the third media data is media data that meets a preset condition.

[0078] The media database includes multiple third media data. The media database can be constructed based on historical and artificial prior knowledge and contains rich media data that users are interested in. The included third media data is diverse and can meet the needs of different users.

[0079] When media data is first produced, its consumption behavior cannot be obtained. Static features can be used to mine similar media data. Static features are inherent characteristics of media data that do not change over time.

[0080] Each first media data in the first media data set is matched with the third media data in the media database, and the first media data in the first media data set that matches the third media data in the media database is obtained, and these first media data are determined as first target media data.

[0081] In an exemplary embodiment, determining the first target media data in the first media data set that matches the third media data in the media database includes: performing feature extraction on the first media data in the first media data set to obtain a first media data feature; searching in a retrieval library based on the first media data feature, wherein the retrieval library includes a third media data feature corresponding to the third media data; if a third media data feature that matches the first media data feature exists in the retrieval library, determining that the first media data is the first target media data.

[0082] An existing self-supervised model or a strongly supervised model (such as a model for extracting music features, image features, text features, etc.) can be used to extract static features of media data, that is, the self-supervised model or the strongly supervised model is used to extract features of each first media data in the first media data set to obtain the first media data feature of each first media data, and the first media data feature is retrieved in a retrieval library. If a third media data feature matching the first media data feature exists in the retrieval library, the first media data corresponding to the first media data feature is determined as the first target media data; if the third media data feature matching the first media data feature does not exist in the retrieval library, the first media data corresponding to the first media data feature is discarded.

[0083] By extracting the first media data features of the first media data in the first media data set and matching the first media data features with the features of the third media data in the media database, it is ensured that the first media data similar to the third media data is determined. Since the third media data in the media database is the media data that the user is interested in, the recalled first media data can also meet the user's needs. Moreover, the third media data in the media database is oriented to most users and has diversity, so the recalled first media data also has diversity.

[0084] In an exemplary embodiment, a search is performed in a retrieval library based on the first media data feature, including: determining a first similarity between the first media data feature and each third media data feature in the retrieval library; if there is a third media data feature whose first similarity is greater than or equal to a first similarity threshold, determining that there is a third media data feature in the retrieval library that matches the first media data feature.

[0085] Calculate the first similarity between the first media data feature and each third media data feature in the retrieval library. If there is a third media data feature in the retrieval library whose first similarity is greater than or equal to the first similarity threshold, determine that the first media data feature matches the third media data feature, that is, there is a third media data feature in the retrieval library that matches the first media data feature. If there is no third media data feature in the retrieval library whose first similarity is greater than or equal to the first similarity threshold, determine that there is no third media data feature in the retrieval library that matches the first media data feature. By determining whether there is a third media data feature in the retrieval library that matches the first media data feature based on whether the first similarity between the first media data feature and the third media data feature in the retrieval library is greater than the first similarity threshold, the accuracy of media data matching can be improved, and it can be ensured that the determined first target media data is similar to the third media data in the media database.

[0086] In step S13, based on the user behavior data, the second target media data having the target dynamic indicator characteristics in the second media data set is obtained, and the target dynamic indicator characteristics are used to characterize the dynamic indicator characteristics corresponding to the obtained second target media data, and the dynamic indicator characteristics are characteristics associated with the user behavior data.

[0087] There may be multiple target dynamic indicator features to determine second target media data having multiple dynamic indicator features to meet the needs of different users. Dynamic features are features that change over time.

[0088] Based on the user behavior data of each second media data in the second media data set, a dynamic indicator feature of each second media data is determined, and second media data having a target dynamic indicator feature is determined as second target media data. The target dynamic indicator feature may be, for example, media data that is frequently downloaded or shared by users.

[0089] In an exemplary embodiment, based on the user behavior data, obtaining the second target media data having the target dynamic indicator feature in the second media data set includes: obtaining the user behavior data of the second media data in the second media data set, and obtaining the target dynamic indicator feature vector corresponding to the target dynamic indicator feature; determining the dynamic indicator feature vector of the second media data based on the user behavior data of the second media data; determining the second similarity between the dynamic indicator feature vector and the target dynamic indicator feature vector; if the second similarity is greater than or equal to a second similarity threshold, determining that the second media data is the second target media data.

[0090] First, the user behavior data corresponding to the media data that is put into the current media data platform after production can be collected, and the machine learning model can be trained based on the various consumption dynamic indicator features of these user behavior data on the media data platform, so that the machine learning model can obtain dynamic indicator feature vectors that can distinguish different dynamic indicator features. After the machine learning model training is completed, the dynamic indicator feature vectors corresponding to the media data with different dynamic indicator features can be determined, that is, the dynamic indicator feature vectors corresponding to different dynamic indicator features can be obtained, and can also be used to determine the dynamic indicator feature vectors of the second media data. The dynamic indicator feature vectors corresponding to different dynamic indicator features can be used to match the dynamic indicator feature vectors of the second media data.

[0091] Each second media data item in the second media data set carries user behavior data, i.e., has dynamic indicator characteristics. The second media data is produced and put into the current media data platform, generating user behavior data. The user behavior data corresponding to each second media data item can then be obtained and input into a machine learning model. The machine learning model then generates a dynamic indicator feature vector corresponding to the second media data item. A target dynamic indicator feature vector corresponding to the target dynamic indicator characteristic is obtained, and a second similarity between the dynamic indicator feature vector of the second media data and the target dynamic indicator feature vector is calculated. If the second similarity is greater than or equal to a second similarity threshold, the second media data item is determined to be the second target media data item. If the second similarity is less than the second similarity threshold, the second media data item is discarded.

[0092] By determining the dynamic indicator feature vector of the second media data based on the user behavior data of the second media data, and calculating the second similarity with the target dynamic indicator feature vector, if the second similarity is greater than or equal to the second similarity threshold, the second media data is determined as the second target media data, which can ensure that the determined second target media data has specific target dynamic indicator characteristics, thereby improving the accuracy of the determined data.

[0093] In step S14, the first target media data, the second target media data, and the third media data are determined as a target data set.

[0094] The first target media data and the second target media data are retrieved based on different characteristics, providing a diverse set of data to meet the needs of different users. The first target media data, the second target media data, and the third media data in the media database are collectively determined as a target data set. This target data set can include newly uploaded first target media data to ensure timeliness, and second target media data with target dynamic indicator characteristics to meet user needs.

[0095] In an exemplary embodiment, the first target media data, the second target media data, and the third media data are determined as a target data set, including: sending the first target media data and the second target media data to a manual review system and obtaining a manual review result; adding the first target media data and the second target media data whose manual review result is passed to the media database; and determining the first target media data and the second target media data, as well as the third media data, which have passed the review in the media database, as the target data set.

[0096] The determined first target media data and second target media data may contain abnormal data, in which case the abnormal data can be eliminated through manual review. After obtaining the first target media data and the second target media data, the first target media data and the second target media data are sent to the manual review system. The manual review system displays the first target media data and the second target media data, and obtains the manual review result. The manual review result for the abnormal first target media data and the second target media data is failure to pass the review, so that the abnormal data can be eliminated, and the first target media data and the second target media data with a manual review result of passing the review are added to the media database to update the media database, ensure the timeliness of the media data in the media database, and determine the first target media data and the second target media data that have passed the review in the media database, as well as the third media data, as the target data set. When delivering media data to other platforms, the media data to be delivered can be obtained from the target data set and delivered to other platforms to promote the current media data platform.

[0097] Since static features have a strong dependency on the media data in the media database, manual secondary verification must be performed before the first target media data and the second target media data enter the media database to eliminate some abnormal data. That is, the first target media data and the second target media data are manually verified and added to the media database only after the verification passes, so as to ensure the cleanliness of the entire media database and avoid the introduction of abnormal data.

[0098] The media data determination method provided by this exemplary embodiment determines the first target media data in the first media data set that matches the third media data in the media database by acquiring the first media data set and the second media data set, and acquires the second target media data in the second media data set having the target dynamic indicator characteristics based on the user behavior data of the second media data, and determines the first target media data and the second target media data and the third media data in the media database as the target data set. Since the characteristics of the media data itself are used to determine the first target media data, and the dynamic characteristics of the media data are used to determine the second target media data, the richness of the determined media data is ensured, and the problem of a single data source for determining the media data is solved.

[0099] Based on the above embodiment, before searching in the retrieval library based on the first media data feature, the method further includes: extracting features of the third media data in the media database respectively to obtain third media data features corresponding to each third media data; and constructing the retrieval library based on the third media data features corresponding to each third media data.

[0100] Before using the search library to search for features of the first media data, a search library must be constructed based on the media database. An existing self-supervised model or a strongly supervised model can be used to extract features from each piece of third media data in the media database, obtaining third media data features corresponding to each piece of third media data. The third media data features corresponding to each piece of third media data are then constructed into a search library, which can then be used to retrieve third media data that matches the first media data. The existing self-supervised model or strongly supervised model is the same model used to extract features from the first media data, ensuring accurate data matching.

[0101] By performing feature extraction on the third media data in the media database in advance and constructing a retrieval library, the efficiency of determining the media data can be guaranteed.

[0102] On the basis of the above technical solution, after determining the first target media data, the second target media data and the third media data as a target data set, the method further includes: displaying the target data set on a target platform interface.

[0103] The target platform may be a platform other than the current media data platform.

[0104] After determining the target data set, the media data in the target data set can be displayed on the target platform interface, for example, it can be delivered to the target platform at a scheduled time. By displaying the target data set on the target platform interface, the target data set can be displayed on more target platforms, expanding the data display platform.

[0105] Figure 2 FIG. 1 is a flow chart showing a method for determining media data according to an exemplary embodiment. Figure 2 As shown, the media data determination method is used in electronic devices such as servers, and includes the following steps.

[0106] In step S21, a first media data set and a second media data set are obtained, where the first media data set includes a plurality of first media data received within a preset time, and the second media data set includes a plurality of second media data, where the second media data carries user behavior data.

[0107] In step S22, feature extraction is performed on the first media data in the first media data set to obtain first media data features.

[0108] In step S23, a search is performed in a search library according to the first media data feature, wherein the search library includes the third media data feature corresponding to the third media data.

[0109] Among them, according to the first media data feature, searching in the retrieval library includes: determining a first similarity between the first media data feature and each third media data feature in the retrieval library; if there is a third media data feature whose first similarity is greater than or equal to a first similarity threshold, determining that there is a third media data feature in the retrieval library that matches the first media data feature.

[0110] Among them, before searching in the retrieval library according to the first media data feature, it also includes: extracting features of the third media data in the media database respectively to obtain the third media data features corresponding to each third media data; and constructing the retrieval library according to the third media data features corresponding to each third media data.

[0111] In step S24, if a third media data feature matching the first media data feature exists in the search library, the first media data is determined to be the first target media data.

[0112] In step S25, user behavior data of the second media data in the second media data set is obtained, and a target dynamic indicator feature vector corresponding to the target dynamic indicator feature is obtained.

[0113] In step S26 , a dynamic indicator feature vector of the second media data is determined based on the user behavior data of the second media data.

[0114] In step S27 , a second similarity between the dynamic index feature vector and the target dynamic index feature vector is determined.

[0115] In step S28 , if the second similarity is greater than or equal to a second similarity threshold, the second media data is determined to be the second target media data.

[0116] In step S29, the first target media data and the second target media data are sent to a manual review system, and a manual review result is obtained;

[0117] In step S210 , the first target media data and the second target media data whose manual review results are passed are added to the media database.

[0118] In step S211 , the first target media data and the second target media data that have passed the review in the media database, as well as the third media data, are determined as the target data set.

[0119] In step S212, the target data set is displayed on the target platform interface.

[0120] The specific content of each step is the same as that of the above embodiment and will not be repeated here.

[0121] The media data determination method provided by this exemplary embodiment uses the characteristics of the media data itself to determine the first target media data and uses the dynamic characteristics of the media data to determine the second target media data, thereby ensuring the richness of the determined media data and solving the problem of a single data source for determining the media data. Moreover, the first media data that does not have user behavior data can also be determined and recalled in a timely manner, thereby improving the timeliness of the recalled media data and improving the efficiency of media data recall.

[0122] Figure 3 Schematic diagram of the implementation of the media data determination method in the embodiment of the present disclosure. Figure 3As shown, when implementing the media data determination method, it can be implemented by a static feature media data determination module, a dynamic feature media data determination module and a manual review module. First, a media database is constructed based on historical and artificial prior knowledge, and the features of the third media data in the media database are extracted through the existing model and a retrieval library is constructed. In the static feature media data determination module, the features of the newly uploaded first media data are extracted using the same model to obtain the first media data features, and the search is performed in the retrieval library. A relatively confident empirical value (such as a first similarity threshold) is used as a judgment condition to determine whether the newly uploaded first media data is similar to the third media data of interest. After the similarity is determined, manual verification is performed. After passing the manual verification, the data can be added to the media database, and the first media data features are added to the retrieval library. In this way, the media data in the media database can be iterated cyclically. In the dynamic feature media data determination module, a machine learning model is first used to determine the dynamic indicator feature vector corresponding to the historical user behavior data of the media data, and a historical consumption database is constructed. For the second media data with user behavior data, the machine learning model is used to determine the dynamic indicator feature vector of the second media data. This is matched with the target dynamic indicator feature vector in the historical consumption database to recall the second target media data with the target dynamic indicator feature. After the second target media data passes manual verification, it is added to the media database. In this way, the media database contains both newly uploaded media data and historical media data with certain target dynamic indicator features, and can be continuously iterated.

[0123] The present disclosure adds a dynamic feature recall source on the basis of using feature similarity, and increases the richness of the recall data in conjunction with the manual review mechanism, which can solve the problem of single recall data due to a single recall link of feature similarity. In addition, the data recalled through dynamic features itself is a reflection of the dynamic consumption indicators of media data, and its richness and effectiveness are greatly improved compared with static features, and it can mine media data of interest efficiently, quickly and in a timely manner.

[0124] Figure 4 FIG. 1 is a block diagram of a device for determining media data according to an exemplary embodiment. Figure 4 The device includes a data set acquisition module 41, a first target data determination module 42, a second target data acquisition module 43 and a target data set determination module 44.

[0125] The data set acquisition module 41 is configured to execute acquisition of a first media data set and a second media data set, wherein the first media data set includes a plurality of first media data received within a preset time, and the second media data set includes a plurality of second media data, wherein the second media data carries user behavior data;

[0126] The first target data determination module 42 is configured to determine first target media data in the first media data set that matches third media data in the media database, where the third media data is media data that meets a preset condition;

[0127] The second target data acquisition module 43 is configured to acquire, based on the user behavior data, second target media data having a target dynamic indicator feature in the second media data set, wherein the target dynamic indicator feature is used to represent a dynamic indicator feature corresponding to the acquired second target media data, and the dynamic indicator feature is a feature associated with the user behavior data;

[0128] The target data set determining module 44 is configured to determine the first target media data, the second target media data, and the third media data as a target data set.

[0129] Optionally, the first target data determination module includes:

[0130] a feature extraction unit, configured to perform feature extraction on the first media data in the first media data set to obtain first media data features;

[0131] a retrieval unit configured to perform a search in a retrieval library based on the first media data feature, wherein the retrieval library includes a third media data feature corresponding to the third media data;

[0132] The first target data determining unit is configured to determine that the first media data is the first target media data if a third media data feature matching the first media data feature exists in the search library.

[0133] Optionally, the retrieval unit is configured to execute:

[0134] determining a first similarity between the first media data feature and each third media data feature in the search library;

[0135] If there is a third media data feature whose first similarity is greater than or equal to a first similarity threshold, it is determined that there is a third media data feature matching the first media data feature in the search library.

[0136] Optionally, the device further includes:

[0137] a seed data feature extraction module configured to perform feature extraction on each of the third media data in the media database to obtain a third media data feature corresponding to each of the third media data;

[0138] The retrieval library construction module is configured to construct the retrieval library according to the third media data feature corresponding to each third media data.

[0139] Optionally, the second target data acquisition module includes:

[0140] A user behavior data acquisition unit configured to acquire user behavior data of the second media data in the second media data set, and acquire a target dynamic indicator feature vector corresponding to the target dynamic indicator feature;

[0141] a dynamic feature vector determining unit configured to determine a dynamic indicator feature vector of the second media data based on user behavior data of the second media data;

[0142] a similarity determination unit configured to determine a second similarity between the dynamic indicator feature vector and the target dynamic indicator feature vector;

[0143] The second target data determining unit is configured to determine that the second media data is the second target media data if the second similarity is greater than or equal to a second similarity threshold.

[0144] Optionally, the recalled data set determination module includes:

[0145] a manual review unit, configured to send the first target media data and the second target media data to a manual review system and obtain a manual review result;

[0146] A seed database data adding unit is configured to add the first target media data and the second target media data, whose manual review results are approved, to the media database;

[0147] The target data set determining unit is configured to determine the first target media data and the second target media data that have passed the review in the media database, as well as the third media data, as the target data set.

[0148] Regarding the apparatus in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0149] Figure 5 5 is a block diagram of an electronic device according to an exemplary embodiment. For example, the electronic device 500 can be provided as a server. Figure 5The electronic device 500 includes a processing component 522, which further includes one or more processors, and a memory resource represented by a memory 532 for storing instructions, such as applications, that can be executed by the processing component 522. The application stored in the memory 532 may include one or more modules, each corresponding to a set of instructions. In addition, the processing component 522 is configured to execute the instructions to perform the above-mentioned media data determination method.

[0150] The electronic device 500 may further include a power supply component 526 configured to perform power management of the electronic device 500, a wired or wireless network interface 550 configured to connect the electronic device 500 to a network, and an input / output (I / O) interface 558. The electronic device 500 may operate based on an operating system stored in the memory 532, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, or the like.

[0151] In an exemplary embodiment, a computer-readable storage medium including instructions is also provided, such as a memory 532 including instructions. The instructions can be executed by the processing component 522 of the electronic device 500 to implement the above-mentioned media data determination method. Alternatively, the computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0152] In an exemplary embodiment, a computer program product is further provided, including a computer program or computer instructions, wherein the computer program or computer instructions implement the above-mentioned media data determination method when executed by a processor.

[0153] Other embodiments of the present disclosure will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the present disclosure being indicated by the following claims.

[0154] It should be understood that the present invention is not limited to the exact construction described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A method for determining media data, characterized in that: include: Obtaining a first media data set and a second media data set, where the first media data set includes a plurality of first media data received within a preset time, and the second media data set includes a plurality of second media data, where the second media data carries user behavior data; Determining first target media data in the first media data set that matches third media data in a media database, where the third media data is media data that meets a preset condition; obtaining, based on the user behavior data, second target media data having a target dynamic indicator feature from the second media data set, the target dynamic indicator feature being used to characterize a dynamic indicator feature corresponding to the obtained second target media data, the dynamic indicator feature being a feature associated with the user behavior data; The first target media data, the second target media data, and the third media data are determined as a target data set.

2. The method according to claim 1, characterized in that Determining first target media data in the first media data set that matches third media data in the media database includes: performing feature extraction on the first media data in the first media data set to obtain first media data features; Searching a search library according to the first media data feature, wherein the search library includes the third media data feature corresponding to the third media data; If a third media data feature matching the first media data feature exists in the search library, the first media data is determined to be the first target media data.

3. The method according to claim 2, characterized in that Searching a search library according to the first media data feature includes: determining a first similarity between the first media data feature and each third media data feature in the search library; If there is a third media data feature whose first similarity is greater than or equal to a first similarity threshold, it is determined that there is a third media data feature matching the first media data feature in the search library.

4. The method according to claim 2, characterized in that Before searching in a search library according to the first media data feature, the method further includes: performing feature extraction on each of the third media data in the media database to obtain third media data features corresponding to each of the third media data; The search library is constructed according to the third media data feature corresponding to each third media data.

5. The method according to any one of claims 1 to 4, characterized in that Acquiring, according to the user behavior data, second target media data having target dynamic indicator characteristics from the second media data set, includes: Obtaining user behavior data of the second media data in the second media data set, and obtaining a target dynamic indicator feature vector corresponding to the target dynamic indicator feature; determining a dynamic indicator feature vector of the second media data based on user behavior data of the second media data; Determining a second similarity between the dynamic indicator feature vector and the target dynamic indicator feature vector; If the second similarity is greater than or equal to a second similarity threshold, the second media data is determined to be the second target media data.

6. The method according to any one of claims 1 to 4, characterized in that Determining the first target media data, the second target media data, and the third media data as a target data set includes: Sending the first target media data and the second target media data to a manual review system, and obtaining a manual review result; adding the first target media data and the second target media data whose manual review results are approved to the media database; The first target media data and the second target media data that have passed the review in the media database, as well as the third media data, are determined as the target data set.

7. The method according to any one of claims 1 to 4, characterized in that After determining the first target media data, the second target media data, and the third media data as a target data set, the method further includes: The target data set is displayed on the target platform interface.

8. A media data determination device, characterized in that: include: a data set acquisition module configured to acquire a first media data set and a second media data set, wherein the first media data set includes a plurality of first media data received within a preset time period, and the second media data set includes a plurality of second media data, wherein the second media data carries user behavior data; a first target data determining module configured to determine first target media data in the first media data set that matches third media data in the media database, where the third media data is media data that meets a preset condition; a second target data acquisition module configured to acquire, from the second media data set, second target media data having a target dynamic indicator feature based on the user behavior data, wherein the target dynamic indicator feature is used to represent a dynamic indicator feature corresponding to the acquired second target media data, the dynamic indicator feature being a feature associated with the user behavior data; The target data set determining module is configured to determine the first target media data, the second target media data, and the third media data as a target data set.

9. The device according to claim 8, characterized in that The first target data determination module includes: a feature extraction unit, configured to perform feature extraction on the first media data in the first media data set to obtain first media data features; a retrieval unit configured to perform a search in a retrieval library based on the first media data feature, wherein the retrieval library includes a third media data feature corresponding to the third media data; The first target data determining unit is configured to determine that the first media data is the first target media data if a third media data feature matching the first media data feature exists in the search library.

10. The device according to claim 9, characterized in that The retrieval unit is configured to perform: determining a first similarity between the first media data feature and each third media data feature in the search library; If there is a third media data feature whose first similarity is greater than or equal to a first similarity threshold, it is determined that there is a third media data feature matching the first media data feature in the search library.

11. The device according to claim 9, characterized in that The device further comprises: a seed data feature extraction module configured to perform feature extraction on each of the third media data in the media database to obtain third media data features corresponding to each of the third media data; The retrieval library construction module is configured to construct the retrieval library according to the third media data feature corresponding to each third media data.

12. The device according to any one of claims 8 to 11, characterized in that The second target data acquisition module includes: A user behavior data acquisition unit configured to acquire user behavior data of the second media data in the second media data set, and acquire a target dynamic indicator feature vector corresponding to the target dynamic indicator feature; a dynamic feature vector determining unit configured to determine a dynamic indicator feature vector of the second media data based on user behavior data of the second media data; a similarity determination unit configured to determine a second similarity between the dynamic indicator feature vector and the target dynamic indicator feature vector; The second target data determining unit is configured to determine that the second media data is the second target media data if the second similarity is greater than or equal to a second similarity threshold.

13. The device according to any one of claims 8 to 11, characterized in that The target data set determination module includes: a manual review unit, configured to send the first target media data and the second target media data to a manual review system and obtain a manual review result; A seed database data adding unit is configured to add the first target media data and the second target media data, whose manual review results are approved, to the media database; The target data set determining unit is configured to determine the first target media data and the second target media data that have passed the review in the media database, as well as the third media data, as the target data set.

14. The device according to any one of claims 8 to 11, characterized in that The device further comprises: The data display module is used to display the target data set in the interface of the target application.

15. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the media data determination method according to any one of claims 1 to 7. 16 . A computer-readable storage medium, wherein when instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the media data determination method according to claim 1 .

17. A computer program product comprising a computer program or computer instructions, characterized in that When the computer program or computer instruction is executed by a processor, the media data determination method according to any one of claims 1 to 7 is implemented.

Images