Audio pushing method, device, storage medium and program product

Abstract

The application discloses an audio pushing method and device, a storage medium and a program product, and belongs to the technical field of computers. The method comprises the following steps: performing clustering processing based on feature information corresponding to each audio to obtain a plurality of audio sets; determining a heat value of each audio based on user operation data corresponding to each audio; for each audio set, determining an audio whose heat value satisfies a heat value condition as a target audio, determining the target audio and an audio having the same performer information as the target audio as a reference original audio, and determining an audio other than the reference original audio in the audio set as a reference pirated audio; and performing audio pushing based on the reference original audio and the reference pirated audio in each audio set. According to the application, when performing audio pushing, the reference original audio and the reference pirated audio determined in each audio set can be comprehensively considered for pushing, so that the influence of the pirated audio is reduced, and the quality of the pushed audio is improved.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to an audio push method, device, storage medium, and program product. Background Technology

[0002] The audio stored in music app libraries consists largely of tracks uploaded directly to the apps and tracks scraped from various sources by web crawlers. This inevitably leads to a significant amount of pirated audio in the libraries. To profit from pirated audio, uploaders typically modify the performer information to differentiate it from legitimate tracks, thus exploiting the platform's reward system. However, the existence of pirated audio severely impacts the profits of the original audio companies.

[0003] The various audio push services of music applications do not take into account the impact of pirated audio. For example, on the recommendation side, if pirated audio meets the recommendation logic, it will still be pushed to the user.

[0004] To date, there is still a lack of audio delivery methods that take into account the impact of audio piracy. Summary of the Invention

[0005] This application provides an audio push method that can solve the problem that the prior art does not take into account the problem of pirated audio when pushing audio.

[0006] Firstly, an audio push method is provided, the method comprising:

[0007] Clustering is performed based on the feature information corresponding to each audio to obtain multiple audio sets, where each audio set includes at least one audio.

[0008] Based on the user operation data corresponding to each audio, the popularity value of each audio is determined;

[0009] For each audio set, audio that meets the popularity value condition is identified as target audio, the target audio and audio with the same performer information as the target audio are identified as reference original audio, and audio in the audio set other than the reference original audio are identified as reference pirated audio.

[0010] Audio is pushed based on the original reference audio and the pirated reference audio in each audio set.

[0011] In one possible implementation, the heat value condition is that the heat value is the largest in the audio set.

[0012] In one possible implementation, the user operation data includes at least one of the following: play count, download count, favorite count, and like count.

[0013] In one possible implementation, determining the popularity value of each audio based on the user operation data corresponding to each audio includes:

[0014] Based on the user operation data and popularity prediction model corresponding to each audio, the popularity value of each audio is determined.

[0015] In one possible implementation, determining the popularity value of each audio based on the user operation data corresponding to each audio includes:

[0016] The popularity value of each audio is determined based on each type of user operation data corresponding to each audio and the first weight value corresponding to each type of user operation data.

[0017] In one possible implementation, the method further includes:

[0018] Based on the attribute information and version comment information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined, wherein the attribute information includes at least one of release time and performer level, and the version comment information includes at least one of the number of legitimate comments and the number of pirated comments;

[0019] The audio push based on the reference original audio and reference pirated audio in each audio set includes:

[0020] Audio is pushed based on the original confidence level of the reference original audio and the reference pirated audio in each audio set.

[0021] In one possible implementation, determining the original confidence level of the reference original audio based on the attribute information and version comment information corresponding to the reference original audio includes:

[0022] Based on the attribute information, version comment information, and version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined, wherein the version tag information includes at least one of the number of times it has been identified as the reference original audio and the number of times it has been identified as the reference pirated audio.

[0023] In one possible implementation, determining the original confidence level of the reference original audio based on the attribute information, version comment information, and version tag information corresponding to the reference original audio includes:

[0024] Based on the original confidence prediction model, as well as the attribute information, version comment information, and version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined.

[0025] In one possible implementation, determining the original confidence level of the reference original audio based on the attribute information, version comment information, and version tag information corresponding to the reference original audio includes:

[0026] Based on the various attribute information, version comment information, version tag information, second weight value for each attribute information, third weight value for each version comment information, and fourth weight value for each version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined.

[0027] In one possible implementation, the audio push based on the original confidence level of the reference original audio and the reference pirated audio in each audio set includes:

[0028] Receive recommendation requests sent by the target terminal;

[0029] Determine the initial recommended audio set corresponding to the target terminal;

[0030] When the initial audio recommendation set includes copyrighted target reference original audio, audio that meets any one of the deletion conditions in the audio set corresponding to the target reference original audio is deleted from the initial recommended audio set to obtain a recommended audio set. The deletion conditions include: being a reference pirated audio; and other reference original audio other than the target reference original audio whose corresponding original confidence is less than the original confidence threshold.

[0031] Each audio file in the recommended audio set is pushed to the target terminal.

[0032] In one possible implementation, the audio push based on reference original audio and reference pirated audio in each audio set includes:

[0033] Receive a search request sent by the target terminal, wherein the search request carries the search content;

[0034] An initial set of search audio is determined based on the search content;

[0035] Replace the reference pirated audio in the initial search audio set with the reference original audio corresponding to the reference pirated audio to obtain the reference search audio set;

[0036] The reference search audio set is deduplicated to obtain the search audio set;

[0037] Each audio file in the searched audio set is pushed to the target terminal.

[0038] In one possible implementation, the search content includes at least one of text search information and audio search information.

[0039] Secondly, an audio delivery device is provided, the device comprising:

[0040] The clustering module is used to perform clustering processing based on the feature information corresponding to each audio to obtain multiple audio sets, wherein each audio set includes at least one audio.

[0041] The first determining module is used to determine the popularity value of each audio based on the user operation data corresponding to each audio.

[0042] The second determining module is used to, for each audio set, determine the audio whose popularity value meets the popularity value condition as the target audio, determine the target audio and the audio with the same performer information as the target audio as the reference original audio, and determine the audio in the audio set other than the reference original audio as the reference pirated audio.

[0043] The push module is used to push audio based on the original reference audio and the pirated reference audio in each audio set.

[0044] In one possible implementation, the heat value condition is that the heat value is the largest in the audio set.

[0045] In one possible implementation, the user operation data includes at least one of the following: play count, download count, favorite count, and like count.

[0046] In one possible implementation, the first determining module is configured to:

[0047] Based on the user operation data and popularity prediction model corresponding to each audio, the popularity value of each audio is determined.

[0048] In one possible implementation, the first determining module is configured to:

[0049] The popularity value of each audio is determined based on each type of user operation data corresponding to each audio and the first weight value corresponding to each type of user operation data.

[0050] In one possible implementation, the device further includes a third determining module, configured to:

[0051] Based on the attribute information and version comment information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined, wherein the attribute information includes at least one of release time and performer level, and the version comment information includes at least one of the number of legitimate comments and the number of pirated comments;

[0052] The push module is used for:

[0053] Audio is pushed based on the original confidence level of the reference original audio and the reference pirated audio in each audio set.

[0054] In one possible implementation, the third determining module is configured to:

[0055] Based on the attribute information, version comment information, and version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined, wherein the version tag information includes at least one of the number of times it has been identified as the reference original audio and the number of times it has been identified as the reference pirated audio.

[0056] In one possible implementation, the third determining module is configured to:

[0057] Based on the original confidence prediction model, as well as the attribute information, version comment information, and version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined.

[0058] In one possible implementation, the third determining module is configured to:

[0059] Based on the various attribute information, version comment information, version tag information, second weight value for each attribute information, third weight value for each version comment information, and fourth weight value for each version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined.

[0060] In one possible implementation, the push module is used to:

[0061] Receive recommendation requests sent by the target terminal;

[0062] Determine the initial recommended audio set corresponding to the target terminal;

[0063] When the initial audio recommendation set includes copyrighted target reference original audio, audio that meets any one of the deletion conditions in the audio set corresponding to the target reference original audio is deleted from the initial recommended audio set to obtain a recommended audio set. The deletion conditions include: being a reference pirated audio; and other reference original audio other than the target reference original audio whose corresponding original confidence is less than the original confidence threshold.

[0064] Each audio file in the recommended audio set is pushed to the target terminal.

[0065] In one possible implementation, the push module is used to:

[0066] Receive a search request sent by the target terminal, wherein the search request carries the search content;

[0067] An initial set of search audio is determined based on the search content;

[0068] Replace the reference pirated audio in the initial search audio set with the reference original audio corresponding to the reference pirated audio to obtain the reference search audio set;

[0069] The reference search audio set is deduplicated to obtain the search audio set;

[0070] Each audio file in the searched audio set is pushed to the target terminal.

[0071] In one possible implementation, the search content includes at least one of text search information and audio search information.

[0072] Thirdly, a computer device is provided, the computer device including a processor and a memory, the memory storing at least one instruction, the instruction being loaded and executed by the processor to implement the operation performed by the audio push method.

[0073] Fourthly, a computer-readable storage medium is provided, wherein at least one instruction is stored in the storage medium, the instruction being loaded and executed by a processor to implement the operations performed by the audio push method.

[0074] Fifthly, a computer program product is provided, the computer program product including at least one instruction, the at least one instruction being loaded and executed by a processor to implement the operations performed by the audio push method.

[0075] The beneficial effects of the technical solution provided in this application embodiment are as follows: The solution mentioned in this application embodiment can first perform clustering processing on each audio based on the feature information corresponding to each audio, thereby obtaining multiple audio sets. Then, based on the user operation data corresponding to each audio, the popularity value of each audio is determined. Audios in each audio set whose popularity value meets the popularity value condition are determined as target audios. Then, the target audios and audios in the audio sets that have the same performer information as the target audios are determined as reference original audios. Audios in the audio sets other than reference original audios are determined as reference pirated audios. Finally, when pushing audio, the reference original audios and reference pirated audios determined in each audio set can be comprehensively considered for pushing, thereby reducing the impact of pirated audios and improving the quality of pushed audios. Attached Figure Description

[0076] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0077] Figure 1 This is a flowchart of an audio push method provided in an embodiment of this application;

[0078] Figure 2 This is a flowchart of an audio push method provided in an embodiment of this application;

[0079] Figure 3 This is a flowchart of an audio push method provided in an embodiment of this application;

[0080] Figure 4 This is a flowchart of an audio push method provided in an embodiment of this application;

[0081] Figure 5 This is a schematic diagram of the structure of an audio push device provided in an embodiment of this application;

[0082] Figure 6 This is a structural block diagram of a terminal provided in an embodiment of this application;

[0083] Figure 7 This is a structural block diagram of a server provided in an embodiment of this application. Detailed Implementation

[0084] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0085] This application provides an audio push method, which can be implemented by a computer device. The computer device can be a terminal or a server, and the terminal can be a desktop computer, laptop computer, tablet computer, mobile phone, etc.

[0086] Computer equipment may include processors, memory, communication components, etc.

[0087] The processor can be a central processing unit (CPU). The processor can be used to read instructions and process data. For example, it can cluster audio files based on feature information to obtain multiple audio sets, determine the popularity value of each audio file based on the user operation information corresponding to each audio file, determine the reference original audio file in each audio set, determine the reference pirated audio file in each audio set, and push audio based on the reference original audio file and the reference pirated audio file in each audio set, etc.

[0088] The memory can be various types of volatile or non-volatile memory, such as solid-state drives (SSDs) and dynamic random access memory (DRAM). Memory can be used for data storage, for example, storing audio data for each audio track, storing data for multiple determined audio sets, storing user operation data and performer information corresponding to each audio track, storing the determined popularity values ​​for each audio track, storing reference original audio and reference pirated audio for each determined audio set, storing intermediate data generated during audio delivery, and so on.

[0089] Communication components can be wired network connectors, wireless fidelity (WiFi) modules, Bluetooth modules, cellular communication modules, etc. These components can be used to transmit data with other devices.

[0090] Figure 1 and Figure 2 This is a flowchart of an audio push method provided in an embodiment of this application. Figure 1 and Figure 2 As shown, this embodiment includes:

[0091] 101. Clustering is performed based on the feature information corresponding to each audio to obtain multiple audio sets.

[0092] In implementation, features can be extracted from each audio file stored in the audio database to obtain the feature information corresponding to each audio file, and then the feature information can be stored in the audio database.

[0093] When it's necessary to distinguish between original and pirated audio, the feature information corresponding to each audio file can be retrieved from an audio database. Then, clustering is performed on this feature information to obtain multiple classes, each representing an audio set, and each audio set includes at least one audio file. Using this method, highly similar or even identical audio files can be clustered to obtain audio sets corresponding to different audio files. For example, for a song M, there are original and cover versions. The original version corresponds to both original and pirated versions, and the cover versions also correspond to both original and pirated versions. Based on the feature information corresponding to each audio file, we can obtain the audio set corresponding to the original version and the audio set corresponding to each cover version.

[0094] When performing clustering, the similarity between the feature information corresponding to each pair of audio files can be calculated. Then, a set of at least one audio file with a similarity greater than or equal to the similarity threshold is determined as an audio set. In one possible implementation, since the audio data of the original audio and the pirated audio are the same or extremely similar, the similarity threshold can be set to a high value. For example, the similarity threshold can be set to a value such as 0.95. This application embodiment does not limit this.

[0095] Of course, the clustering method can be any other reasonable method, and the embodiments of this application are not limited thereto.

[0096] The aforementioned audio-related feature information can be obtained using any reasonable method. For example, features can be extracted from an audio feature extraction model trained on the audio input to obtain the output feature information corresponding to the audio. Alternatively, the audio can be segmented into frames, windowed, and then a Fast Fourier Transform can be performed on each frame to extract amplitude information and calculate the power spectrum. Then, 33 non-overlapping frequency bands distributed in the 300-2000 Hz range are selected from the power spectral density to calculate 32 sub-fingerprints for each frame. Based on the 32 sub-fingerprints corresponding to multiple audio frames, the audio fingerprint corresponding to the audio is determined, and this audio fingerprint is the audio-related feature information. Of course, the feature information can also be information in different forms obtained by other methods, and this application embodiment does not limit this.

[0097] 102. Determine the popularity value of each audio based on the user operation data corresponding to each audio.

[0098] In practice, after identifying multiple audio sets, the original audio and pirated audio can be initially identified for each audio set, i.e., the original audio and the pirated audio can be referenced.

[0099] Since original audio is usually distributed through more abundant and legitimate channels, user activity data for original and pirated audio will differ significantly after a period of time. This can be used as one of the factors to determine whether an audio is original or pirated.

[0100] User operation data can include at least one of the following: number of plays, number of downloads, number of favorites, and number of likes. Generally, original audio will have more plays, downloads, favorites, or likes than pirated audio.

[0101] The popularity score of each audio can be calculated based on user action data corresponding to each audio. There are several possible calculation methods:

[0102] The first method is to directly use any type of user action data as the audio's popularity score. For example, the number of plays or downloads could be used as the audio's popularity score, and so on.

[0103] The second method involves determining the popularity value of each audio based on user action data and a popularity prediction model corresponding to each audio. In practice, when calculating the popularity value of an audio, the user action data corresponding to that audio can be input into the trained popularity prediction model to obtain the output popularity value of that audio.

[0104] The third method determines the popularity value of each audio based on each type of user operation data and the first weight value corresponding to each type of user operation data. In implementation, a corresponding first weight value can be set for each type of user operation data. When calculating the popularity value of an audio, each type of user operation data corresponding to that audio can be multiplied by the corresponding first weight value, and then the weighted user operation data can be added together to obtain the popularity value of that audio.

[0105] The methods for calculating the heat value of audio are not limited to the above-mentioned methods, but can be any other reasonable methods. This application does not limit these methods.

[0106] 103. For each audio set, the audio that meets the popularity value condition is identified as the target audio, the target audio and the audio with the same performer information as the target audio are identified as the original reference audio, and the audio in the audio set other than the original reference audio are identified as the pirated reference audio.

[0107] In practice, since the original audio usually has a higher popularity value, after determining the popularity value of each audio, the audio in the audio set that meets the popularity value condition can be initially identified as the reference original audio and used as the target audio. Since the performer information of the pirated audio is different from that of the original audio, all audio in the audio set that has the same performer information as the target audio can be identified as the reference original audio. Then, the remaining audio in the audio set, excluding the reference original audio, can be identified as the reference pirated audio.

[0108] In one possible implementation, the heat value condition can be that the audio with the highest heat value in the audio set is selected as the target audio.

[0109] The following are examples:

[0110] The audio set includes audio A, audio B, audio C, audio D, and audio F, with performer information of apple, pear, peach, apple, and pear, respectively. The calculated popularity values ​​of audio A, audio B, audio C, audio D, and audio F are 100, 150, 80, 20, and 300, respectively. Therefore, audio F with the highest popularity value can be identified as the target audio in the audio set. Then, audio F and other audios with performer information of pear (i.e., audio B) can be identified as the original reference audio. The remaining audio A, audio C, and audio D can be identified as the pirated reference audio.

[0111] 104. Based on the original reference audio and the pirated reference audio in each audio set, audio is pushed out.

[0112] In practice, since the identified original audio source is likely to be the original audio source, and the identified pirated audio source may be pirated audio source, when pushing audio, we can consider whether to push the audio source based on the original audio source or the pirated audio source, thereby minimizing the impact of pirated audio and improving the quality of the pushed audio source.

[0113] In one possible implementation, when pushing audio, in addition to considering whether the pushed audio is based on the original or pirated audio, the original confidence score (i.e., the probability that the reference original audio is the original audio) of each reference original audio can be calculated. The push is then made by comprehensively considering the reference original audio and its corresponding original confidence score. The corresponding method can be as follows:

[0114] Based on the attribute information and version comment information corresponding to the original audio, the confidence level of the original audio is determined. The attribute information includes at least one of release time and performer level, and the version comment information includes at least one of the number of legitimate comments and the number of pirated comments. When pushing audio, the audio is pushed based on the confidence level of the original audio and the number of pirated audio in each audio set.

[0115] In practice, the confidence level of each reference original audio can be determined based on its corresponding attribute information and version comment information. Specifically, the number of pirated comments is the number of comments containing the word "pirated" or a semantically similar word among all the comment data for that reference original audio, while the number of legitimate comments is the number of comments containing the word "legitimate" or a semantically similar word among all the comment data for that reference original audio. Generally, for a reference original audio, the older its release date, the higher the performer's level, the fewer the number of pirated comments, and the more the number of legitimate comments, the greater the likelihood that it is the original audio, and thus the higher its confidence level of authenticity.

[0116] When pushing content, in addition to considering whether each audio is based on the original or a pirated version, the system can also use the confidence level of the original version for each reference audio to push the reference audio that is more likely to be the original to the user. This reduces the impact of pirated audio and improves the quality of the push notification.

[0117] In one possible implementation, since the audio database is frequently updated with new audio files or deleted audio files, and users may also modify the performer information of audio files, steps 101-103 can be performed periodically to improve the accuracy of each audio push. Each process determines the version information of each audio file in the audio database, identifying the original reference audio based on its popularity value and performer information, and identifying any audio files other than the original reference audio as pirated reference audio. Furthermore, after determining the version information of each audio file, it can be marked, recording the number of times each audio file is identified as the original reference audio and the number of times it is identified as a pirated reference audio, using this as one of the factors to consider when determining the confidence level of the original version.

[0118] Alternatively, steps 101-104 can be iterated repeatedly. That is, after each audio push in step 104, step 101 is processed again to redetermine each audio set and the reference original audio and reference pirated audio within it. Then, the number of times each audio is identified as the reference original audio and the number of times it is identified as the reference pirated audio are recorded, and these are used as one of the factors to consider when determining the confidence level of the original.

[0119] Correspondingly, the confidence level of the original audio can be determined based on the attribute information, version comment information, and version tagging information corresponding to the original audio. The version tagging information includes at least one of the number of times it was identified as a reference to the original audio and the number of times it was identified as a reference to pirated audio.

[0120] Generally, the more times an audio is identified as a reference to the original source and the fewer times it is identified as a reference to pirated audio, the greater the likelihood that the referenced audio is the original source, and the greater its confidence level as the original source.

[0121] There are several methods for determining the confidence level of a reference original audio based on its corresponding attribute information, version comment information, and version tag information. Two of these methods are shown below:

[0122] The first method: Based on the original confidence prediction model, as well as the attribute information, version comment information, and version tag information corresponding to the original audio, determine the original confidence corresponding to the reference original audio.

[0123] In practice, the attribute information, version comment information, and version tag information corresponding to the original audio can be input into the trained original confidence prediction model to obtain the output original confidence corresponding to the reference original audio.

[0124] The second method is to determine the original confidence level of the reference original audio based on multiple attribute information, version comment information, version tag information, a second weight value for each attribute information, a third weight value for each version comment information, and a fourth weight value for each version tag information.

[0125] In implementation, a second weight value can be assigned to each type of attribute information, a third weight value to each version of comment information, and a fourth weight value to each version of tag information. When determining the confidence level of the original audio version, the corresponding weight value can be multiplied by the original value and then added together to obtain the confidence level of the original audio version.

[0126] Of course, other reasonable methods can be used to determine the confidence level of the original version, and this application does not limit this method.

[0127] In one possible implementation, the confidence level of the reference pirated audio can be determined based on the attribute information corresponding to the reference pirated audio. The method for determining this is similar to the method for determining the confidence level of the original version, and will not be repeated here.

[0128] In one possible implementation, the confidence level of the original audio file corresponding to each reference original audio file and the confidence level of the pirated audio file corresponding to each reference pirated audio file can be considered together to make subsequent audio pushes.

[0129] Below, we will provide a more detailed introduction to the two aspects of audio push notifications: recommendation-based audio push and search-based audio push.

[0130] like Figure 3 As shown, the recommended methods for side-audio push are as follows:

[0131] 301. Receive the recommendation request sent by the target terminal.

[0132] In practice, when the target terminal opens the application, the target terminal sends a recommendation request to the computer device, which then receives the recommendation request, which may include an account identifier.

[0133] 302. Determine the initial recommended audio set corresponding to the target terminal.

[0134] In practice, after receiving a recommendation request, the computer device can determine the initial set of recommended audio based on the account identifier carried in the recommendation request. For example, it can first obtain the historical operation information corresponding to the account identifier based on the account identifier, and then determine multiple audios that the user is interested in corresponding to the account identifier based on the historical operation information, thus forming the initial set of recommended audio.

[0135] 303. When the initial audio recommendation set includes copyrighted target reference original audio, delete any audio from the audio set corresponding to the target reference original audio that meets any deletion condition in the initial recommended audio set to obtain the recommended audio set.

[0136] In implementation, the initial recommended audio may include multiple audio sets, or a subset of audio from each set. Since copyrighted original audio is highly likely to be the original source material, if a copyrighted original audio (i.e., the target original audio) is present in the initial recommended audio, it indicates that a relatively stable original source resource already exists for that audio. In this case, other audio in the corresponding audio set that meets the deletion criteria—that is, other audio with the potential for piracy—can be removed. This increases the proportion of original audio in the recommended audio set, thus improving the recommendation quality.

[0137] If there is no copyrighted original audio reference in the initial set of recommended audio, a stable recommended audio cannot be determined. In subsequent deletion operations, there is a possibility that multiple audios in that set will be deleted. Therefore, if there is no copyrighted original audio reference in the initial set of recommended audio, no deletion operation is required.

[0138] The deletion criteria include ① referencing pirated audio, and ② other reference original audio besides the target reference original audio whose corresponding original confidence level is less than the original confidence threshold.

[0139] Regarding point ①, in the audio set corresponding to the target reference original audio, since the reference pirated audio may be pirated, adhering to the principle of rigor, since there are relatively stable original recommended resources (i.e., the target reference original audio), these reference pirated audio can be deleted from the initial audio recommendation set.

[0140] Regarding point ②, in the audio set corresponding to the target reference original audio, other reference original audio besides the target reference original audio can be filtered based on the original confidence level. Other reference original audio with an original confidence level lower than the original confidence level threshold can be deleted from the initial audio recommendation set.

[0141] In the initial audio recommendation set, audio that meets either of the above two deletion conditions is deleted, thus obtaining a recommended audio set. In this recommended audio set, some audio with a high probability of piracy is deleted, while stable original recommended resources are ensured, thereby improving the recommendation quality and bringing a good user experience.

[0142] 304. Push each audio file from the recommended audio set to the target terminal.

[0143] In practice, after the recommended audio set is determined, the computer device can send each audio in the recommended audio set to the target terminal. After receiving each audio in the recommended audio set, the target terminal can display its information on the recommendation page of the application's homepage.

[0144] like Figure 4 As shown, the search-side audio push method can be as follows:

[0145] 401. Receive the search request sent by the target terminal.

[0146] The search request contains the search terms.

[0147] In practice, users can open the application on the target terminal, enter their search terms, and click the "Search" button. This will trigger the target terminal to send a search request to the computer device, which will carry the search terms entered by the user.

[0148] The search content can include at least one of text search information and audio search information. When a user enters text in the application's search bar to search for audio, that text becomes the text search information; when a user uses the song recognition function in the application, the entered information becomes the audio search information. Of course, the search content can also include both text search information and audio search information; this embodiment of the application does not limit this.

[0149] 402. Determine the initial search audio set based on the search content.

[0150] In practice, after receiving a search request, the computer device can determine the initial set of search audio based on the search content carried in the search request.

[0151] For example, if a user enters the text search information "a flower", the computer device will retrieve multiple audio clips containing the three characters "a flower" in the corresponding text information, thus obtaining the initial search audio set.

[0152] For example, if the user inputs audio search information as audio segment A, the computer device can perform similarity detection between audio segment A and various audio segments stored in the audio database, and then determine multiple audio segments with similarity greater than a similarity threshold as the initial search audio set.

[0153] 403. Replace the reference pirated audio in the initial search audio set with the reference original audio corresponding to the reference pirated audio to obtain the reference search audio set.

[0154] In practice, the initial search audio set may include reference original audio and reference pirated audio from multiple audio sets. In this case, the reference pirated audio in the initial search audio set can be replaced with other reference original audio in the same audio set as the reference pirated audio, thus obtaining a reference search audio set. In this way, all audio in the reference search audio set is reference original audio.

[0155] When replacing the pirated audio with the corresponding original audio, if the audio set includes multiple original audios, one original audio can be randomly selected to replace the pirated audio. Alternatively, the selection can be based on popularity, i.e., the original audio with the highest popularity can be selected to replace the pirated audio. Or, the pirated audio can be replaced with all the original audios in the corresponding audio set, etc. This application does not limit the selection method.

[0156] 404. Perform deduplication on the reference search audio set to obtain the search audio set.

[0157] In practice, since the initial search audio set may contain both the original reference audio and the pirated reference audio, replacing the pirated reference audio with the corresponding original reference audio may result in duplicate original reference audio. In this case, the audio in the search audio set can be deduplicated to obtain a search audio set without duplicate audio.

[0158] 405. Push each audio file in the search audio collection to the target terminal.

[0159] In practice, after determining the search audio set, the computer device can send each audio in the search audio set to the target terminal. After receiving each audio in the search audio set, the target terminal can display its information on the application's search page.

[0160] The audio push in this application embodiment is not limited to the two types mentioned above. When performing other audio push processing, subsequent audio push can also be performed based on the reference original audio in each audio set and the original confidence level corresponding to the reference original audio.

[0161] In one possible implementation, the database can be organized based on the original reference audio and the pirated reference audio in each determined audio set. For example, reference pirated audio that is not copyrighted and has a piracy confidence level less than the piracy confidence threshold can be directly deleted from the database.

[0162] All of the above-mentioned optional technical solutions can be combined in any way to form the optional embodiments of this application, and will not be described in detail here.

[0163] The scheme mentioned in this application embodiment can first cluster the audio based on the feature information corresponding to each audio to obtain multiple audio sets. Then, based on the user operation data corresponding to each audio, the popularity value of each audio is determined. Audios in each audio set whose popularity value meets the popularity value condition are determined as target audios. Then, the target audios and audios in the audio sets that have the same performer information as the target audios are determined as reference original audios. Other audios in the audio sets besides the reference original audios are determined as reference pirated audios. Finally, when pushing audio, the reference original audios and reference pirated audios determined in each audio set can be comprehensively considered for push, thereby reducing the impact of pirated audios and improving the quality of pushed audios.

[0164] This application provides an audio push device, which can be the computer device described in the above embodiments, such as... Figure 5 As shown, the device includes:

[0165] Clustering module 510 is used to perform clustering processing based on the feature information corresponding to each audio to obtain multiple audio sets, wherein each audio set includes at least one audio.

[0166] The first determining module 520 is used to determine the popularity value of each audio based on the user operation data corresponding to each audio.

[0167] The second determining module 530 is used to, for each audio set, determine the audio whose popularity value meets the popularity value condition as the target audio, determine the target audio and the audio with the same performer information as the target audio as the reference original audio, and determine the audio in the audio set other than the reference original audio as the reference pirated audio.

[0168] The push module 540 is used to push audio based on the reference original audio and reference pirated audio in each audio set.

[0169] In one possible implementation, the heat value condition is that the heat value is the largest in the audio set.

[0170] In one possible implementation, the user operation data includes at least one of the following: play count, download count, favorite count, and like count.

[0171] In one possible implementation, the first determining module 520 is configured to:

[0172] Based on the user operation data and popularity prediction model corresponding to each audio, the popularity value of each audio is determined.

[0173] In one possible implementation, the first determining module 520 is configured to:

[0174] The popularity value of each audio is determined based on each type of user operation data corresponding to each audio and the first weight value corresponding to each type of user operation data.

[0175] In one possible implementation, the device further includes a third determining module, configured to:

[0176] Based on the attribute information and version comment information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined, wherein the attribute information includes at least one of release time and performer level, and the version comment information includes at least one of the number of legitimate comments and the number of pirated comments;

[0177] The push module 540 is used for:

[0178] Audio is pushed based on the original confidence level of the reference original audio and the reference pirated audio in each audio set.

[0179] In one possible implementation, the third determining module is configured to:

[0180] Based on the attribute information, version comment information, and version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined, wherein the version tag information includes at least one of the number of times it has been identified as the reference original audio and the number of times it has been identified as the reference pirated audio.

[0181] In one possible implementation, the third determining module is configured to:

[0182] Based on the original confidence prediction model, as well as the attribute information, version comment information, and version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined.

[0183] In one possible implementation, the third determining module is configured to:

[0184] Based on the various attribute information, version comment information, version tag information, second weight value for each attribute information, third weight value for each version comment information, and fourth weight value for each version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined.

[0185] In one possible implementation, the push module 540 is used for:

[0186] Receive recommendation requests sent by the target terminal;

[0187] Determine the initial recommended audio set corresponding to the target terminal;

[0188] When the initial audio recommendation set includes copyrighted target reference original audio, audio that meets any one of the deletion conditions in the audio set corresponding to the target reference original audio is deleted from the initial recommended audio set to obtain a recommended audio set. The deletion conditions include: being a reference pirated audio; and other reference original audio other than the target reference original audio whose corresponding original confidence is less than the original confidence threshold.

[0189] Each audio file in the recommended audio set is pushed to the target terminal.

[0190] In one possible implementation, the push module 540 is used for:

[0191] Receive a search request sent by the target terminal, wherein the search request carries the search content;

[0192] An initial set of search audio is determined based on the search content;

[0193] Replace the reference pirated audio in the initial search audio set with the reference original audio corresponding to the reference pirated audio to obtain the reference search audio set;

[0194] The reference search audio set is deduplicated to obtain the search audio set;

[0195] Each audio file in the searched audio set is pushed to the target terminal.

[0196] In one possible implementation, the search content includes at least one of text search information and audio search information.

[0197] It should be noted that the audio push device provided in the above embodiments is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the audio push device and the audio push method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0198] Figure 6A structural block diagram of a terminal 600 provided in an exemplary embodiment of this application is shown. The terminal can be a computer device as described in the above embodiments. The terminal 600 can be a smartphone, tablet computer, MP3 player (Moving Picture Experts Group Audio Layer III), MP4 player (Moving Picture Experts Group Audio Layer IV), laptop computer, or desktop computer. The terminal 600 may also be referred to as a user device, portable terminal, laptop terminal, desktop terminal, or other names.

[0199] Typically, terminal 600 includes a processor 601 and a memory 602.

[0200] Processor 601 may include one or more processing cores, such as a quad-core processor, an octa-core processor, etc. Processor 601 may be implemented using at least one hardware form selected from DSP (digital signal processing), FPGA (field-programmable gate array), and PLA (programmable logic array). Processor 601 may also include a main processor and a coprocessor. The main processor, also known as a CPU (central processing unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, processor 601 may integrate a GPU (graphics processing unit), which is responsible for rendering and drawing the content to be displayed on the screen. In some embodiments, processor 601 may also include an AI (artificial intelligence) processor, which is used to handle computational operations related to machine learning.

[0201] The memory 602 may include one or more computer-readable storage media, which may be non-transitory. The memory 602 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the non-transitory computer-readable storage media in the memory 602 are used to store at least one instruction, which is executed by the processor 601 to implement the audio push method provided in the method embodiments of this application.

[0202] In some embodiments, the terminal 600 may also optionally include a peripheral device interface 603 and at least one peripheral device. The processor 601, memory 602, and peripheral device interface 603 can be connected via a bus or signal line. Each peripheral device can be connected to the peripheral device interface 603 via a bus, signal line, or circuit board. Specifically, the peripheral device includes at least one of the following: a radio frequency circuit 604, a display screen 605, a camera 606, an audio circuit 607, a positioning component 608, and a power supply 609.

[0203] Peripheral interface 603 can be used to connect at least one I / O (input / output) related peripheral device to processor 601 and memory 602. In some embodiments, processor 601, memory 602 and peripheral interface 603 are integrated on the same chip or circuit board; in some other embodiments, any one or two of processor 601, memory 602 and peripheral interface 603 can be implemented on separate chips or circuit boards, which is not limited in this embodiment.

[0204] The radio frequency (RF) circuit 604 is used to receive and transmit RF (radio frequency) signals, also known as electromagnetic signals. The RF circuit 604 communicates with communication networks and other communication devices via electromagnetic signals. The RF circuit 604 converts electrical signals into electromagnetic signals for transmission, or converts received electromagnetic signals back into electrical signals. Optionally, the RF circuit 604 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a user identity module card, etc. The RF circuit 604 can communicate with other terminals through at least one wireless communication protocol. This wireless communication protocol includes, but is not limited to: metropolitan area networks (MANs), various generations of mobile communication networks (2G, 3G, 4G, and 5G), wireless local area networks (WLANs), and / or WiFi (wireless fidelity) networks. In some embodiments, the RF circuit 604 may also include circuitry related to NFC (near field communication), which is not limited in this application.

[0205] Display screen 605 is used to display a user interface (UI). This UI may include graphics, text, icons, videos, and any combination thereof. When display screen 605 is a touch display screen, it also has the ability to collect touch signals on or above its surface. These touch signals can be input as control signals to processor 601 for processing. In this case, display screen 605 can also be used to provide virtual buttons and / or a virtual keyboard, also known as soft buttons and / or a soft keyboard. In some embodiments, there may be one display screen 605, which serves as the front panel of terminal 600; in other embodiments, there may be at least two display screens, respectively disposed on different surfaces of terminal 600 or in a folded design; in still other embodiments, display screen 605 may be a flexible display screen, disposed on a curved or folded surface of terminal 600. Furthermore, display screen 605 may be configured as a non-rectangular, irregular shape, i.e., a non-rectangular screen. Display screen 605 may be made of materials such as LCD (liquid crystal display) or OLED (organic light-emitting diode).

[0206] The camera assembly 606 is used to acquire images or videos. Optionally, the camera assembly 606 includes a front-facing camera and a rear-facing camera. Typically, the front-facing camera is located on the front panel of the terminal, and the rear-facing camera is located on the back of the terminal. In some embodiments, there are at least two rear-facing cameras, which are any one of a main camera, a depth-sensing camera, a wide-angle camera, and a telephoto camera, to achieve background blurring by fusion of the main camera and the depth-sensing camera, panoramic shooting by fusion of the main camera and the wide-angle camera, VR (virtual reality) shooting, or other fusion shooting functions. In some embodiments, the camera assembly 606 may also include a flash. The flash can be a single-color temperature flash or a dual-color temperature flash. A dual-color temperature flash refers to a combination of a warm-light flash and a cool-light flash, which can be used for light compensation at different color temperatures.

[0207] The audio circuit 607 may include a microphone and a speaker. The microphone is used to collect sound waves from the user and the environment, converting the sound waves into electrical signals that are input to the processor 601 for processing, or input to the radio frequency circuit 604 for voice communication. For stereo sound acquisition or noise reduction purposes, multiple microphones may be used, each located at a different part of the terminal 600. The microphone may also be an array microphone or an omnidirectional microphone. The speaker is used to convert the electrical signals from the processor 601 or the radio frequency circuit 604 into sound waves. The speaker may be a conventional diaphragm speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can convert electrical signals not only into audible sound waves but also into inaudible sound waves for purposes such as distance measurement. In some embodiments, the audio circuit 607 may also include a headphone jack.

[0208] The positioning component 608 is used to determine the current geographic location of the terminal 600 in order to enable navigation or LBS (location-based service). The positioning component 608 can be a positioning component based on GPS (global positioning system), BeiDou system, Grenadin system, or Galileo system.

[0209] Power supply 609 is used to supply power to the various components in terminal 600. Power supply 609 can be AC ​​power, DC power, a disposable battery, or a rechargeable battery. When power supply 609 includes a rechargeable battery, the rechargeable battery can support wired or wireless charging. The rechargeable battery can also be used to support fast charging technology.

[0210] In some embodiments, the terminal 600 further includes one or more sensors 610. The one or more sensors 610 include, but are not limited to: an accelerometer 611, a gyroscope 612, a pressure sensor 613, a fingerprint sensor 614, an optical sensor 615, and a proximity sensor 616.

[0211] Accelerometer 611 can detect the magnitude of acceleration along the three coordinate axes of a coordinate system established by terminal 600. For example, accelerometer 611 can be used to detect the components of gravitational acceleration along the three coordinate axes. Processor 601 can control display screen 605 to display the user interface in either a landscape or portrait view based on the gravitational acceleration signal acquired by accelerometer 611. Accelerometer 611 can also be used for games or for acquiring user motion data.

[0212] The gyroscope sensor 612 can detect the orientation and rotation angle of the terminal 600. The gyroscope sensor 612, in conjunction with the accelerometer sensor 611, can collect 3D motion data from the user on the terminal 600. Based on the data collected by the gyroscope sensor 612, the processor 601 can perform the following functions: motion sensing (e.g., changing the UI based on the user's tilt), image stabilization during shooting, game control, and inertial navigation.

[0213] The pressure sensor 613 can be disposed on the side bezel of the terminal 600 and / or on the lower layer of the display screen 605. When the pressure sensor 613 is disposed on the side bezel of the terminal 600, it can detect the user's grip signal on the terminal 600, and the processor 601 can perform left / right hand recognition or quick operation based on the grip signal collected by the pressure sensor 613. When the pressure sensor 613 is disposed on the lower layer of the display screen 605, the processor 601 can control the operable controls on the UI interface based on the user's pressure operation on the display screen 605. The operable controls include at least one of button controls, scroll bar controls, icon controls, and menu controls.

[0214] The fingerprint sensor 614 is used to collect a user's fingerprint. The processor 601 identifies the user based on the fingerprint collected by the fingerprint sensor 614, or vice versa. When the user's identity is identified as trusted, the processor 601 authorizes the user to perform relevant sensitive operations, including unlocking the screen, viewing encrypted information, downloading software, making payments, and changing settings. The fingerprint sensor 614 can be located on the front, back, or side of the terminal 600. When the terminal 600 has physical buttons or a manufacturer's logo, the fingerprint sensor 614 can be integrated with the physical buttons or manufacturer's logo.

[0215] An optical sensor 615 is used to collect ambient light intensity. In one embodiment, the processor 601 can control the display brightness of the display screen 605 based on the ambient light intensity collected by the optical sensor 615. Specifically, when the ambient light intensity is high, the display brightness of the display screen 605 is increased; when the ambient light intensity is low, the display brightness of the display screen 605 is decreased. In another embodiment, the processor 601 can also dynamically adjust the shooting parameters of the camera assembly 606 based on the ambient light intensity collected by the optical sensor 615.

[0216] The proximity sensor 616, also known as a distance sensor, is typically mounted on the front panel of the terminal 600. The proximity sensor 616 is used to detect the distance between the user and the front of the terminal 600. In one embodiment, when the proximity sensor 616 detects that the distance between the user and the front of the terminal 600 is gradually decreasing, the processor 601 controls the display screen 605 to switch from a screen-on state to a screen-off state; when the proximity sensor 616 detects that the distance between the user and the front of the terminal 600 is gradually increasing, the processor 601 controls the display screen 605 to switch from a screen-off state to a screen-on state.

[0217] Those skilled in the art will understand that Figure 6 The structure shown does not constitute a limitation on terminal 600, and may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0218] Figure 7 This is a schematic diagram of a server structure provided in an embodiment of this application. The server 700 can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 701 and one or more memories 702. The memories 702 store at least one instruction, which is loaded and executed by the processors 701 to implement the methods provided in the above-described method embodiments. Of course, the server may also have wired or wireless network interfaces, a keyboard, and input / output interfaces for input and output. The server may also include other components for implementing device functions, which will not be elaborated upon here.

[0219] In an exemplary embodiment, a computer-readable storage medium is also provided, such as a memory including instructions that can be executed by a processor in a terminal to complete the audio push method described above. This computer-readable storage medium can be non-transitory. For example, the computer-readable storage medium can be ROM (read-only memory), RAM (random access memory), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0220] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0221] It should be noted that all information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.), and signals (including but not limited to signals transmitted between user terminals and other devices) involved in this application have been authorized by the user or fully authorized by all parties, and the collection, use, and processing of related data must comply with the relevant laws, regulations, and standards of the relevant countries and regions. For example, the "user operation data" involved in this application was obtained under full authorization.

[0222] The above description is merely an optional embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. An audio push method, characterized in that, The method includes: Clustering is performed based on the feature information corresponding to each audio to obtain multiple audio sets, where each audio set includes at least one audio. Based on the user operation data corresponding to each audio, the popularity value of each audio is determined; For each audio set, audio that meets the popularity value condition is identified as target audio, the target audio and audio with the same performer information as the target audio are identified as reference original audio, and audio in the audio set other than the reference original audio are identified as reference pirated audio. Based on the attribute information and version comment information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined. The attribute information includes at least one of release time and performer level. The version comment information includes at least one of the number of legitimate comments and the number of pirated comments. The number of pirated comments is the number of comments in all comment data of the reference original audio that include the word "pirated" or a word semantically similar to "pirated." The number of legitimate comments is the number of comments in all comment data of the reference original audio that include the word "legitimate" or a word semantically similar to "legitimate." Audio push is performed based on the reference original audio and reference pirated audio in each audio set, including: audio push based on the original confidence level corresponding to the reference original audio and the reference pirated audio in each audio set.

2. The method according to claim 1, characterized in that, The heat value condition is that the heat value is the highest in the audio set.

3. The method according to claim 2, characterized in that, The user operation data includes at least one of the following: number of plays, number of downloads, number of favorites, and number of likes.

4. The method according to claim 3, characterized in that, The process of determining the popularity value of each audio based on the user operation data corresponding to each audio includes: Based on the user operation data and popularity prediction model corresponding to each audio, the popularity value of each audio is determined.

5. The method according to claim 3, characterized in that, The process of determining the popularity value of each audio based on the user operation data corresponding to each audio includes: The popularity value of each audio is determined based on each type of user operation data corresponding to each audio and the first weight value corresponding to each type of user operation data.

6. The method according to claim 1, characterized in that, The step of determining the original confidence level of the reference original audio based on the attribute information and version comment information corresponding to the reference original audio includes: Based on the attribute information, version comment information, and version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined, wherein the version tag information includes at least one of the number of times it has been identified as the reference original audio and the number of times it has been identified as the reference pirated audio.

7. The method according to claim 6, characterized in that, The step of determining the original confidence level of the reference original audio based on the attribute information, version comment information, and version tag information corresponding to the reference original audio includes: Based on the original confidence prediction model, as well as the attribute information, version comment information, and version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined.

8. The method according to claim 6, characterized in that, The step of determining the original confidence level of the reference original audio based on the attribute information, version comment information, and version tag information corresponding to the reference original audio includes: Based on the various attribute information, version comment information, version tag information, second weight value for each attribute information, third weight value for each version comment information, and fourth weight value for each version tag information corresponding to the reference original audio, the original confidence level corresponding to the reference original audio is determined.

9. The method according to claim 1, characterized in that, The audio push based on the original confidence level of the reference original audio and the reference pirated audio in each audio set includes: Receive recommendation requests sent by the target terminal; Determine the initial recommended audio set corresponding to the target terminal; When the initial recommended audio set includes copyrighted target reference original audio, audio that meets any one of the deletion conditions in the audio set corresponding to the target reference original audio is deleted from the initial recommended audio set to obtain a recommended audio set. The deletion conditions include: being a reference pirated audio; and other reference original audio other than the target reference original audio whose corresponding original confidence is less than the original confidence threshold. Each audio file in the recommended audio set is pushed to the target terminal.

10. The method according to claim 1, characterized in that, The audio push based on the reference original audio and reference pirated audio in each audio set also includes: Receive a search request sent by the target terminal, wherein the search request carries the search content; An initial set of search audio is determined based on the search content; Replace the reference pirated audio in the initial search audio set with the reference original audio corresponding to the reference pirated audio to obtain the reference search audio set; The reference search audio set is deduplicated to obtain the search audio set; Each audio file in the searched audio set is pushed to the target terminal.

11. The method according to claim 10, characterized in that, The search content includes at least one of text search information and audio search information.

12. A computer device, characterized in that, The computer device includes a processor and a memory, the memory storing at least one instruction, which is loaded and executed by the processor to perform the operation of the audio push method as described in any one of claims 1 to 11.

13. A computer-readable storage medium, characterized in that, The storage medium stores at least one instruction, which is loaded and executed by a processor to perform the operation of the audio push method as described in any one of claims 1 to 11.

14. A computer program product, characterized in that, The computer program product includes at least one instruction, which is loaded and executed by a processor to perform the operation of the audio push method as described in any one of claims 1 to 11.

Images