JP7061679B2 - A method and system for predicting the playing length of a song based on the composition of the playlist - Google Patents

Description

The following description relates to a technique for providing a streaming-based music service.

With the development of digital technology, it has become possible to generate, store and manage sound sources in digital form. This makes it easier to create, distribute, and manage sound sources, and it has become possible to easily view them using various terminals such as PCs, notebook PCs, tablets, and smartphones.

While the methods for listening to music are diversifying, for example, music can be listened to by downloading and using it via a communication network or receiving it in real time by a streaming service via the Internet. It became so.

As an example, Patent Document 1 (publication date: February 14, 2014) discloses a playlist-based music broadcasting service method and the apparatus for providing a user-customized music broadcasting service based on a user's playlist. Has been done.

Korean Publication No. 10-2014-0019123 Gazette

The purpose is to predict the playback length, which is the probability that the corresponding song will be played to the end for each song, based on the composition of the songs in the playlist (playlist).

Another object is to provide a smooth playback environment by effectively performing prefetching for each song based on the playback length of each song included in the playlist.

It is possible to reduce unnecessary data fetches based on the playback length of each song in the playlist, and the other purpose is to efficiently use the resources consumed throughout the service.

A method performed by a computer system, wherein the computer system includes at least one processor, wherein the at least one processor manages song composition information and playback history information for each playlist. When the at least one processor receives a playback request for a specific playlist from an electronic device, the corresponding song is played for each song in the specific playlist using the song composition information and the playback history information. Provided is a method including a step of predicting a reproduction length which is a probability of being performed.

According to one aspect, the management step is for each of the playlists with respect to the song composition information including the playback order of the songs and the total playback length of each song, and the total playback length of each song. It may include a step of managing the playback history information including the actual playback length indicating the ratio of the actual playback by each user.

According to another aspect, the predictive stage uses the song composition information to acquire a playlist corresponding to the song composition of the specific playlist, and uses the playback history information of the acquired playlist. The step of calculating the expected playback length for each song in the specific playlist may be included.

Further, according to another aspect, the acquisition step may acquire a playlist composed of songs that are the same as or similar to the songs included in the specific playlist, based on the similarity between the songs.

According to another aspect, the acquisition stage is a stage in which the sound source data of the corresponding song is learned by the deep learning learning model to generate a unique sound source feature for each song, and the song using the sound source feature. It may include a step of calculating the similarity between the songs and a step of acquiring a playlist composed of songs that are the same as or similar to the songs included in the specific playlist based on the similarity between the songs.

According to another aspect, the prediction stage includes a stage of acquiring a playlist corresponding to the composition of a song in the specific playlist by using the song composition information, and the specific stage of the acquired playlist. It may include a step of calculating an expected playback length for each song in the specific playlist by a method of averaging the actual playback lengths of all users for each song in the playlist and the corresponding song.

According to another aspect, the prediction stage includes a stage of acquiring a playlist corresponding to the song composition of the specific playlist by using the song composition information, and the specific play of the acquired playlist. A step of calculating an expected playback length for each song in the specific playlist by a method of averaging the actual playback lengths of the users of the electronic device for each song in the list and the corresponding song may be included.

According to another aspect, the management stage adds additional information to each song in the playlist, including at least one of the time, weather, and position at the time the song was played. The step of further including the step of managing together with the information, the step of predicting is the step of acquiring the playlist corresponding to the song composition of the specific playlist by using the song composition information, and the stage of acquiring the playlist. It may include a step of calculating an expected playback length for each song of the specific playlist by using the playlist playback history information that matches the additional information at the time when the specific playlist is requested.

According to another aspect, in the method, the at least one processor distributes the resources required for song playback according to the playback length of each song of the specific playlist, and the method is for playing the specific playlist. It may further include a step of providing the sound source data of the above to the electronic device.

According to yet another aspect, the provided step may perform prefetching for each song in the particular playlist in units corresponding to the predicted playback length of the song.

Provided is a non-temporary computer-readable recording medium, characterized in that a program for causing a computer to execute the method is recorded.

A computer system comprising a memory and at least one processor connected to the memory and configured to execute computer-readable instructions contained in the memory, wherein the at least one processor is each playlist. When a playback request for a specific playlist is received from the information management unit that manages the song composition information and the playback history information, the specific playlist is used by using the song composition information and the playback history information. Provided is a computer system including a playback length predicting unit that predicts a playback length, which is the probability that the corresponding song is played for each song in the playlist.

It is a figure which showed the example of the network environment in one Embodiment of this invention. It is a block diagram for demonstrating the internal structure of the electronic device and the server in one Embodiment of this invention. It is a block diagram which showed the example of the component which the processor of a server can include in one Embodiment of this invention. It is a flowchart which showed the example of the method which a server can execute in one Embodiment of this invention. It is a figure which showed the example of the management table which contains the music composition information of a playlist in one Embodiment of this invention. It is a figure which showed the example of the management table which contains the play history information of a playlist in one Embodiment of this invention. It is a flowchart which showed an example of the process of calculating the expected play length for each song of a playlist based on the degree of similarity between songs in one Embodiment of this invention. It is an exemplary diagram for demonstrating the process of acquiring the playlist corresponding to the song composition of a specific playlist in one embodiment of the present invention. It is an exemplary diagram for demonstrating the process of calculating the expected reproduction length for each music of a specific playlist in one Embodiment of this invention. It is an exemplary figure for demonstrating the prefetch process by the expected play length for each song of a specific playlist in one Embodiment of this invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

An embodiment of the present invention relates to a technique for providing a streaming-based music service.

Embodiments, including the matters specifically disclosed herein, can predict the playing length of a song based on a playlist structure, thereby efficiency, resource savings, cost savings, quality of service, convenience. It is possible to achieve considerable advantages in terms of sex and the like.

FIG. 1 is a diagram showing an example of a network environment according to an embodiment of the present invention. The network environment of FIG. 1 shows an example including a plurality of electronic devices 110, 120, 130, 140, a plurality of servers 150, 160, and a network 170. Such FIG. 1 is merely an example for explaining the invention, and the number of electronic devices and the number of servers are not limited as in FIG.

The plurality of electronic devices 110, 120, 130, 140 may be fixed terminals or mobile terminals realized by a computer system. Examples of a plurality of electronic devices 110, 120, 130, 140 include smartphones, mobile phones, navigation systems, PCs (personal computers), notebook PCs, digital broadcasting terminals, PDAs (Personal Digital Assistants), and PMPs (Tablet Multimedia Players). ), Tablets, game consoles, wearable devices, IoT (internet of things) devices, VR (visual reality) devices, AR (agmented reality) devices, and the like. As an example, FIG. 1 shows a smartphone as an example of the electronic device 110, but in the embodiment of the present invention, the electronic device 110 substantially utilizes a wireless or wired communication method, and another via the network 170. It may mean one of a variety of physical computer systems capable of communicating with the electronic devices 120, 130, 140 and / or the servers 150, 160.

The communication method is not limited, and not only a communication method using a communication network that can be included in the network 170 (for example, a mobile communication network, a wired Internet, a wireless Internet, a broadcasting network, a satellite network, etc.), but also a device. Short-range wireless communication between may be included. For example, the network 170 includes a PAN (personal area network), a LAN (local area network), a CAN (campus area network), a MAN (metropolitan area network), a WAN (wise Internet) network, etc. It may include any one or more of the networks. Further, network 170 may include, but is limited to, any one or more of network topologies, including bus networks, star networks, ring networks, mesh networks, star-bus networks, tree or hierarchical networks, and the like. Will not be done.

Each of the servers 150, 160 is realized by one or more computer devices that communicate with a plurality of electronic devices 110, 120, 130, 140 via a network 170 to provide instructions, codes, files, contents, services, and the like. good. For example, the server 150 may be a system that provides the first service to a plurality of electronic devices 110, 120, 130, 140 connected via the network 170, and the server 160 may also be a plurality of systems connected via the network 170. It may be a system that provides a second service to electronic devices 110, 120, 130, 140. As a more specific example, the server 150 is a service (as an example, a music service) aimed at by the application through an application which is a computer program installed and executed in a plurality of electronic devices 110, 120, 130, 140. Etc.) may be provided to a plurality of electronic devices 110, 120, 130, 140 as a first service. As another example, the server 160 may provide a service for distributing the above-mentioned application installation and execution files to a plurality of electronic devices 110, 120, 130, 140 as a second service.

FIG. 2 is a block diagram for explaining an internal configuration of an electronic device and a server according to an embodiment of the present invention. FIG. 2 describes the internal configuration of the electronic device 110 and the internal configuration of the server 150 as examples for the electronic device. Further, the other electronic devices 120, 130, 140 and the server 160 may have the same or similar internal configuration as the above-mentioned electronic device 110 or the server 150.

The electronic device 110 and the server 150 may include memories 211, 221s, processors 212, 222, communication modules 213, 223, and input / output interfaces 214, 224. The memory 211 and 221 are non-temporary computer-readable recording media, and are a RAM (random access memory), a ROM (read only memory), a disk drive, an SSD (sold state drive), and a flash memory (flash memory). Permanent mass recording devices such as, etc. may be included. Here, even if a permanent large-capacity recording device such as a ROM, SSD, flash memory, or disk drive is included in the electronic device 110 or the server 150 as another permanent recording device that is separated from the memories 211 and 221. good. In addition, the memory 211 and 221 contain an operating system and at least one program code (for example, a browser installed and executed in the electronic device 110 or an application installed in the electronic device 110 to provide a specific service. Code for etc.) may be recorded. Such software components may be loaded from a computer-readable recording medium separate from the memories 211 and 221. Such other computer-readable recording media may include computer-readable recording media such as floppy (registered trademark) drives, discs, tapes, DVD / CD-ROM drives, and memory cards. In other embodiments, software components may be loaded into memory 211, 221 through communication modules 213, 223, which are not computer readable recording media. For example, at least one program is a computer program installed by a file provided by a file distribution system (eg, server 160 described above) that distributes a developer or application installation file via a network 170 (eg, described above). It may be loaded into the memory 211 or 221 based on the application.

Processors 212 and 222 may be configured to process instructions in a computer program by performing basic arithmetic, logic, and input / output operations. Instructions may be provided to processor 212, 222 by memory 211, 221 or communication modules 213, 223. For example, the processors 212 and 222 may be configured to execute instructions received according to a program code recorded in a recording device such as memories 211 and 221.

The communication modules 213 and 223 may provide a function for the electronic device 110 and the server 150 to communicate with each other via the network 170, and the electronic device 110 and / or the server 150 may provide another electronic device (as an example). , Electronic device 120) or another server (eg, server 160). As an example, a request generated by a processor 212 of an electronic device 110 according to a program code recorded in a recording device such as a memory 211 may be transmitted to a server 150 via a network 170 under the control of a communication module 213. On the contrary, control signals, instructions, contents, files and the like provided under the control of the processor 222 of the server 150 are received by the electronic device 110 through the communication module 213 of the electronic device 110 via the communication module 223 and the network 170. It's okay. For example, control signals, instructions, contents, files, etc. of the server 150 received through the communication module 213 may be transmitted to the processor 212 and the memory 211, and the contents, files, etc. may be further recorded by the electronic device 110. It may be recorded on a medium (permanent recording device described above).

The input / output interface 214 may be a means for an interface with the input / output device 215. For example, the input device may include a device such as a keyboard, mouse, microphone, camera, and the output device may include a device such as a display, speaker, haptic feedback device, and the like. As another example, the input / output interface 214 may be a means for an interface with a device that integrates functions for input and output, such as a touch screen. The input / output device 215 may be composed of an electronic device 110 and one device. Also, the input / output interface 224 of the server 150 may be a means for connecting to the server 150 or for interfacing with a device (not shown) for input or output that the server 150 can include. As a more specific example, a service screen or content configured by using data provided by the server 150 or the electronic device 120 when the processor 212 of the electronic device 110 processes an instruction of a computer program loaded in the memory 211. May be displayed on the display through the input / output interface 214.

Also, in other embodiments, the electronic device 110 and the server 150 may include more components than the components of FIG. However, most prior art components need not be clearly shown in the figure. For example, the electronic device 110 may be realized to include at least a part of the above-mentioned input / output device 215, such as a transceiver, a GPS (Global Positioning System) module, a camera, various sensors, a database, and the like. Other components may be further included. As a more specific example, when the electronic device 110 is a smartphone, an acceleration sensor or gyro sensor, a camera module, various physical buttons, a button using a touch panel, an input / output port, which are generally included in the smartphone, are used. Various components, such as accelerometers for vibration, may be realized to be further included in the electronic device 110.

Hereinafter, a method of predicting the playback length of a song based on a playlist structure and a specific embodiment of the system will be described.

In a typical playlist, songs are played in sequence according to simple rules. Since the resources for playback are consumed even for songs that are likely to be skipped in the middle without being played to the end by the user, such unnecessary resource consumption causes the overall consumption of resources. Service may not be smooth.

It is an object of the present invention to predict the flow of playlist reproduction by a probabilistic model and to effectively allocate the resources required for song reproduction according to the prediction.

FIG. 3 is a block diagram showing an example of components that can be included in the processor of the server in one embodiment of the present invention, and FIG. 4 is a block diagram of what the server executes in one embodiment of the present invention. It is a flowchart which showed the example of the possible method.

The server 150 according to the present embodiment serves as a platform for providing a streaming-based music service to a plurality of electronic devices 110, 120, 130, 140 which are clients (clients). The server 150 may provide music services in conjunction with applications installed on the electronic devices 110, 120, 130, 140.

As shown in FIG. 3, the processor 222 of the server 150 includes an information management unit 310, a reproduction length prediction unit 320, and a data provision unit 330 as components for executing the music providing method according to FIG. good. Depending on the embodiment, the components of processor 222 may be selectively included or excluded from processor 222. Also, depending on the embodiment, the components of processor 222 may be separated or merged to represent the functionality of processor 222.

Such a processor 222 and the components of the processor 222 may control the server 150 to perform steps 410 to 440 included in the music providing method of FIG. For example, the processor 222 and the components of the processor 222 may be implemented to execute an instruction by the code of the operating system included in the memory 221 and the code of at least one program.

Here, the components of the processor 222 may be representations of different functions of the processor 222 that are executed by the processor 222 according to the instructions provided by the program code recorded in the server 150. For example, the information management unit 310 may be used as a functional representation of the processor 222 that controls the server 150 according to the instructions described above so that the server 150 manages the playlist and related information.

At step 410, processor 222 may read the required instructions from memory 221 loaded with instructions associated with control of server 150. In this case, the read instructions may include instructions for controlling the processor 222 to perform steps 420 to 440 described below.

In the stage 420, the information management unit 310 may investigate and aggregate the song composition information and the reproduction history information for each playlist, and record and manage them. A playlist is a collection of tracks composed of at least one piece of music content (songs), and includes a playlist directly composed by the user or shared by others, or a playlist automatically recommended on the service. It's fine. The information management unit 310 may record and manage general information about the song composition of each playlist or playback log data by the user for playlists having a history of consumption by the user in the music service. FIG. 5 is a diagram showing an example of a management table including the song composition information 500 of the playlist_A. As shown in FIG. 5, with respect to the playlist_A configured in the order of song_1-> song_2-> song_3-> song_4, the song composition information 500 is a subject such as a song playback order 501, a theme or title of the song composition. 502, the total playback length 503, the average playback length 504, the playback length 505 for each song, the genre 506, the BPM (beat per minute) 507, and the like may be included. FIG. 6 is a diagram showing an example of a management table including the playback history information 600 of the playlist_A. Suppose all users play songs in the same order for playlist_A. As shown in FIG. 6, for playlist_A configured in the order of song_1-> song_2-> song_3-> song_4, the playback history information 600 includes the song playback order 601 and the overall playback length of each song by each user. It may include an actual reproduction length 602 or the like, which is a reproduction ratio indicating how much the reproduction is actually performed. Further, the information management unit 310 additionally collects the time zone, weather, position, social issue, etc. at the time of song playback as additional information regarding the time when the song is played for each song in playlist_A. It may be managed together with the actual playback length 602 of each user. Such additional information is acquired on the Internet by using the information acquired from the electronic devices 110, 120, 130, 140 of the user who uses the music service, or the information acquired from the electronic devices 110, 120, 130, 140. It may mean a comprehensive set of information.

Referring to FIG. 4 again, in the stage 430, when the playback request for the specific playlist is received from the electronic device 110, the playback length prediction unit 320 together with the song composition information of the playlist managed in the stage 420. The playback history information may be used to predict the playback length of each song according to the song composition of a specific playlist. The reproduction length prediction unit 320 may predict the reproduction flow of a specific playlist requested by the user of the electronic device 110 by a probabilistic model. In the present invention, a model capable of associating the song composition information of each playlist with the reproduction history information may be designed, and pre-calculated data may be secured by using such a model. For example, among the playlists including the song composition in which Track2 is arranged after Track1, 90 seconds of the total playback time of 120 seconds is played by Track2 on average in the playlist of subject A, and 90 seconds in the playlist of subject B is played. On average, Track 2 may play 80 seconds out of a total playback time of 120 seconds. If the user requests the reproduction of a specific playlist, the reproduction length prediction unit 320 may acquire the log data of the existing playlist corresponding to the song composition of the corresponding playlist and share it with the electronic device 110. .. As an example, the playback length prediction unit 320 may use an existing playlist corresponding to the song composition of a specific playlist requested by the electronic device 110 from among playlists having a playback history by the user, based on the similarity between songs. May be searched and the expected playback length may be calculated for each song in a specific playlist by using the playback history information of the corresponding playlist.

FIG. 7 is a flowchart showing an example of a process of calculating an expected reproduction length for each song in a playlist based on the similarity between songs in one embodiment of the present invention.

In step 701, when the sound source data of the corresponding song is input to each song, the reproduction length prediction unit 320 may process the input sound source data in the form of learning data. At this time, the music content may mean all digital data having an audio file format, and includes, for example, MP3 (MPEG AudioLayer-3), WAVE (Waveform AudioFormat), FLAC (Free Lossless AudioCodec), and the like. It's fine. Further, text information about the song may be input to the playback length prediction unit 320 together, but at this time, the text information includes lyrics or meta information such as a singer, a genre, a song name, an album name, and the like. , Further, information such as a hashtag and a query input in connection with song classification and search may be included. Subsequently, the reproduction length prediction unit 320 may express the sound source data in time-frequency by preprocessing. For example, the reproduction length prediction unit 320 may convert the sound source data into time-frequency-magnitude form data such as a mel spectrogram or MFCC (Mel Frequency Cepstral Cofficient). When the text information for each song is input to the reproduction length prediction unit 320, the text information may be preprocessed. As an example, the reproduction length prediction unit 320 may filter meaningless text from the input text information by a language preprocessing device such as a morphological analyzer, an index term extractor, and the like. In other words, the reproduction length prediction unit 320 removes part-speech words such as particles and auxiliary verbs and special symbols (for example,!,?, / Etc.) contained in the text information, and corresponds to an uninflected word or a root. Words may be extracted.

In step 702, the reproduction length prediction unit 320 learns the training data preprocessed for the sound source data by the learning model to generate unique features, and then stores the generated unique features in a database (not shown). ) May be recorded. The reproduction length prediction unit 320 may generate the unique acoustic features of the sound source data itself by using deep learning. As an example, the reproduction length prediction unit 320 may utilize a learning model based on a convolutional neural network (CNN). The reproduction length prediction unit 320 may express the sound source data as a multidimensional real number vector by using the CNN learning model. The CNN learning model may include a sound source data learning layer, and the following steps 1 to 3 may be an example of a process of generating a real number vector corresponding to the sound source data in the sound source data learning layer.

In process 1, the sound source data of the song (for example, an mp3 file) may be converted into time-frequency-magnitude form data such as mel spectrogram or MFCC by preprocessing.

In step 2, a plurality of frequency frames in one or more short time intervals (1 to 10 seconds) may be sampled from the converted sound source data and used as input data for the learning model of the sound source data. For example, the reproduction length prediction unit 320 may sample a plurality of frames and utilize it as an input of a CNN model presented as an example of a music model in the sound source data learning layer. As a result, the CNN model for learning the sound source data becomes a model having the same number of channels as the sampled frames. Alternatively, use each generated frame as a single channel, go through a unique convolution / pooling process for each frame, and then join the feature vectors of each generated frame to use as an input for the fully connected layer. It will be possible.

In step 3, a plurality of convolution and pooling layers that can be included in the sound source data learning layer may be iteratively configured to generate an abstracted feature from the music frame. In convolution, the size of the patch may be variously configured, and the pooling technique is a pooling technique that uses the maximum value (max), a pooling technique that uses the average value (average), and a hybrid pooling technique that combines the two pooling techniques. At least one of a variety of pooling techniques, such as, may be used.

On multiple convolution and pooling layers, there is a full-connected layer for generating acoustic features, and the functions for each layer are sigmoid function, hyperbolic tangent (tanh) function, and ReLU. Various functions such as the (Rectifier Unit) function may be used. After all, one multidimensional real number vector may be generated for the sound source data. For example, the given first sound source data m0 is one multidimensional real number in the output layer of the music learning model, such as the form of x0 = {0.2, -0.1, 0.3, ...}. It may be expressed as a vector.

Further, when the text information is input together for each song, the reproduction length prediction unit 320 may also express the text information as a multidimensional real number vector. As an example, the reproduction length prediction unit 320 may generate a text filtered by preprocessing for text information as a word vector (word vector) by a pre-learned learning model. For example, a word vector may be represented in a numerical multidimensional vector form. For word vector generation, a word appearance frequency histogram, TF (term frequency) / IDF (inverse document fraction), language learning model (for example, word2vec, phase2vec, document2vec, etc.) may be used. For example, for a singer / genre / song title / year such as "Lee Seung Hwan Good Day Ballad 1992", one n-dimensional real number vector v = {0.3, -1.2, 1.2, ...} It can be expressed as. At this time, it is not necessary to fix the text information fields (singer, genre, song title, year, etc.) for the language learning model in this order, and it can be changed to suit the purpose.

The reproduction length prediction unit 320 described above may record and maintain a feature vector of sound source data and a word vector of text information in a database for each song. The feature vector of the sound source data and the word vector of the text information may be constructed in individual databases or may be constructed in one database, respectively. Such a database may be realized by the components included in the server 150, or may exist as an external database built on a separate system that can work with the server 150.

In step 703, the playlist length prediction unit 320 may calculate the similarity between songs by using the unique characteristics of the sound source data recorded in the database for each song, and based on the similarity between songs. , The playlist corresponding to the song composition of the specific playlist requested by the electronic device 110 may be acquired from the playlists having the playback history by the user. As another example, the reproduction length prediction unit 320 can calculate the similarity between songs by using the word vector of the text information in combination with the feature vector of the sound source data. In other words, the reproduction length prediction unit 320 may acquire a playlist composed of songs that are the same as or similar to the songs of the specific playlist requested by the electronic device 110, based on the similarity between the songs.

In the stage 704, the playback length prediction unit 320 uses the playback history information of the playlist acquired in the stage 703 to actually replay the songs constituting the specific playlist and the corresponding songs by a large number of users. The expected playback length for each song in a specific playlist may be calculated by a method of averaging the growth length.

As shown in FIG. 8, when the specific playlist requested by the electronic device 110 is the Playlist_I 800 configured in the order of song_i-> song_iii-> song_iii-> song_iv, the reproduction length predicting unit 320 determines. Songs whose composition is the same as or similar to song_i, song_ii, song_iii, song_iv (song_i', song_i'', ..., song_ii', song_ii'', ..., song_iii', song_iii'', ... Search for playlists (Playlist_1 to N) 810 including', song_iv'', ...). In the playlist (Playlist_1 to N) 810, at least two or more songs in the song structure song_i-> song_iii-> song_iv of Playlist_I800 are continuous (for example, song_i-> song_ii, song_ii-> sion. -> Song_iv) and the corresponding playlist may be included.

Referring to FIG. 9, the reproduction length prediction unit 320 searches for a playlist 810 corresponding to a specific playlist 800 requested by the electronic device 110, and averages the reproduction length for each song of the corresponding playlist 810. You can do it. For example, by a user who requests playback of a particular playlist by averaging the actual play lengths of songs that are the same as or similar to the first song in a particular playlist (song_i, song_i', song_i''). You may predict how much song_i will be played. If the user's playback length based on the playback history information of the playlist 810 that is the same as or similar to the song composition of the specific playlist 800 is 90%, 10%, 20%, or 75% on average, song_i-> song_ii The playback length of each song may be predicted to be 90%, 10%, 20%, and 75% for the Playlist_I800 configured in the order of-> song_iii-> song_iv.

Therefore, the playback length prediction unit 320 averages the playback logs of other users who have consumed playlists having the same or similar song composition for the specific playlist requested by the user, so that the playback length predicting unit 320 can be used for the specific playlist. The playback length can be predicted for each song. At this time, when the user's playback log for the playlist is sufficiently accumulated, the playback length prediction unit 320 uses only the user's playback log instead of the other user's playback log to obtain a specific playlist. It is also possible to personalize and predict the playback length of each song. In addition, the playback length prediction unit 320 uses a playback log having additional information that matches the time zone, weather, position, social issue, etc. at the time when the electronic device 110 requests a specific playlist, so that the user can use the playback length prediction unit 320. It is also possible to predict the playback length for each song in a particular playlist depending on the situation.

Referring to FIG. 4 again, in the stage 440, the data providing unit 330 distributes and specifies the resource corresponding to the playback length for each song predicted in the stage 430 to the specific playlist requested by the electronic device 110. The sound source data for reproducing the playlist may be provided to the electronic device 110. The data providing unit 330 may effectively distribute the resources required for playing a song according to the predicted playing length of each song for a specific playlist, and perform prefetching. For example, as shown in FIG. 10, the playback length of Playlist_I800 configured in the order of song_i-> song_ii-> second_iii-> song_iv based on the playback logs of other users who have consumed playlists having the same or similar composition of songs. When it is predicted that the value is 90%, 10%, 20%, and 75% for each song, the electronic device 110 receives and reproduces the sound source data of song_i in chunks of 9.0 seconds at the same time, and at the same time, song_ii. Is 1.0 second, song_iii is 2.0 seconds, and song_iv is 7.5 seconds. The corresponding sound source data is downloaded in advance to prepare for the next playback. In other words, the data providing unit 330 prefetches more data earlier for a specific playlist 800 requested by the user of the electronic device 110 in the case of a song having a high probability of being played to the end, and to the end. For songs with a low probability of being played, less data is prefetched more slowly.

As described above, according to the embodiment of the present invention, it is possible to predict the playback length, which is the probability that the corresponding song is played to the end for each song, based on the song structure of the playlist, and the song is based on this. By efficiently executing prefetch for each, a smooth playback environment can be provided. Further, according to an embodiment of the present invention, it is possible to reduce unnecessary data fetches for playlists based on the playback length of each song, thereby efficiently utilizing the resources consumed throughout the service. be able to.

The devices described above may be implemented by hardware components, software components, and / or combinations of hardware components and software components. For example, the apparatus and components described in the embodiments include a processor, a controller, an ALU (arithmetic logic unit), a digital signal processor, a microcomputer, an FPGA (field programgable gate array), a PLU (programmable log unit), a microprocessor, and the like. Alternatively, it may be implemented using one or more general purpose computers or special purpose computers, such as various devices capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the OS. The processing device may also respond to the execution of the software, access the data, and record, manipulate, process, and generate the data. For convenience of understanding, one processing device may be described as being used, but one of ordinary skill in the art may include a plurality of processing elements and / or a plurality of types of processing elements. You can understand. For example, the processing device may include multiple processors or one processor and one controller. Also, other processing configurations such as parallel processors are possible.

The software may include computer programs, codes, instructions, or a combination of one or more of these, configuring the processing equipment to operate at will, or instructing the processing equipment independently or collectively. You may do it. The software and / or data is embodied in any type of machine, component, physical device, computer recording medium or device to be interpreted based on the processing device or to provide instructions or data to the processing device. good. The software is distributed on a computer system connected by a network and may be recorded or executed in a distributed state. The software and data may be recorded on one or more computer-readable recording media.

The method according to the embodiment may be realized in the form of program instructions that can be executed by various computer means and recorded on a computer-readable medium. Here, the medium may be a continuous recording of a computer-executable program or a temporary recording for execution or download. Further, the medium may be various recording means or storage means in the form of a combination of a single piece of hardware or a plurality of pieces of hardware, and is not limited to a medium directly connected to a certain computer system, but is distributed over a network. It may exist. Examples of media include hard disks, floppy (registered trademark) disks, magnetic media such as magnetic tapes, optical media such as CD-ROMs and DVDs, optical magnetic media such as floptic discs, and the like. And may include ROM, RAM, flash memory, etc., and may be configured to record program instructions. Other examples of media include recording media or storage media managed by application stores that distribute applications, sites that supply or distribute various other software, servers, and the like.

As described above, the embodiments have been described based on the limited embodiments and drawings, but those skilled in the art will be able to make various modifications and modifications from the above description. For example, the techniques described may be performed in a different order than the methods described, and / or components such as the systems, structures, devices, circuits described may be in a different form than the methods described. Appropriate results can be achieved even if they are combined or combined, and confronted or replaced by other components or equivalents.

Therefore, even if it is a different embodiment, if it is equivalent to the claims, it belongs to the attached claims.

Claims (20)

The way a computer system does
The computer system includes at least one processor.
The method is
A stage in which the at least one processor manages song composition information and playback history information for each playlist.
When the at least one processor receives a playback request for a specific playlist from an electronic device, the corresponding song is played for each song in the specific playlist using the song composition information and the playback history information. A method that includes a step of predicting the play length, which is a probability. The management stage is
For each of the playlists, the song composition information including the playback order of the songs and the total playback length of each song, and the actual ratio of each user to the total playback length of each song. The method according to claim 1, comprising a step of managing playlist information including the playback length of the song. The prediction stage is
The stage of acquiring the playlist corresponding to the song composition of the specific playlist by using the song composition information, and
The method according to claim 1 or 2, comprising the step of calculating the expected playback length for each song of the specific playlist using the playback history information of the acquired playlist. The acquisition stage is
The method according to claim 3, wherein a playlist composed of songs that are the same as or similar to the songs included in the specific playlist is acquired based on the degree of similarity between the songs. The acquisition stage is
For each song, the stage of learning the sound source data of the corresponding song by the deep learning learning model to generate unique sound source features, and
At the stage of calculating the similarity between songs using the sound source characteristics,
The method according to claim 3, comprising the step of acquiring a playlist composed of songs that are the same as or similar to the songs included in the specific playlist based on the similarity between the songs. The prediction stage is
The stage of acquiring the playlist corresponding to the song composition of the specific playlist by using the song composition information, and
A step of calculating an expected playback length for each song of the specific playlist by a method of averaging the actual playback lengths of all users for each song of the specific playlist in the acquired playlist. The method according to claim 2, wherein the method comprises. The prediction stage is
The stage of acquiring the playlist corresponding to the song composition of the specific playlist by using the song composition information, and
The expected playback length of each song of the specific playlist is calculated by a method of averaging the actual playback lengths of the users of the electronic device for the songs corresponding to each song of the specific playlist in the acquired playlist. The method of claim 2, comprising the step of calculation. The management stage is
Further including the step of managing additional information including at least one of the time, weather, and position at the time when the corresponding song is played for each song in the playlist together with the playback history information.
The prediction stage is
The stage of acquiring the playlist corresponding to the song composition of the specific playlist by using the song composition information, and
Among the acquired playlists, the expected playback length for each song of the specific playlist is calculated by using the playlist playback history information that matches the additional information at the time when the specific playlist is requested. The method of claim 2, comprising the steps. The method is
The at least one processor distributes resources necessary for playing a song according to the playing length of each song in the specific playlist, and provides sound source data for playing the specific playlist to the electronic device. The method of any one of claims 1-8, further comprising a step. The above-mentioned stage of provision is
The method of claim 9, wherein for each song in the particular playlist, prefetching is performed in units corresponding to the predicted playback length of the song. A program for causing a computer to execute the method according to any one of claims 1 to 10. It ’s a computer system,
Includes memory and at least one processor connected to said memory and configured to execute computer-readable instructions contained in said memory.
The at least one processor
An information management unit that manages song composition information and playback history information for each playlist,
When a playback request for a specific playlist is received from an electronic device, the playback length is the probability that the song is played for each song in the specific playlist using the song composition information and the playback history information. A computer system that includes a playlist length predictor that predicts the song. The information management department
For each of the playlists, the song composition information including the song playback order and the total playback length of each song, and the actual ratio of each song actually played by each user to the total playback length of each song. The computer system according to claim 12, wherein the playback history information including the playback length is managed. The reproduction length prediction unit is
Using the song composition information, the playlist corresponding to the song composition of the specific playlist is acquired, and the playlist is obtained.
The computer system according to claim 12 or 13, wherein the expected playback length is calculated for each song of the specific playlist by using the playback history information of the acquired playlist. The reproduction length prediction unit is
The computer system according to claim 14, wherein a playlist composed of songs that are the same as or similar to the songs included in the specific playlist is acquired based on the similarity between the songs. The reproduction length prediction unit is
For each song, the sound source data of the corresponding song is learned by a deep learning learning model to generate unique sound source features.
Using the sound source characteristics, the similarity between songs is calculated, and
The computer system according to claim 14, wherein a playlist composed of songs that are the same as or similar to the songs included in the specific playlist is acquired based on the similarity between the songs. The reproduction length prediction unit is
Using the song composition information, the playlist corresponding to the song composition of the specific playlist is acquired, and the playlist is obtained.
To calculate the expected playback length for each song in the specific playlist by a method of averaging the actual playback lengths of all users for the songs corresponding to each song in the specific playlist in the acquired playlist. 13. The computer system according to claim 13. The reproduction length prediction unit is
Using the song composition information, the playlist corresponding to the song composition of the specific playlist is acquired, and the playlist is obtained.
The expected playback length for each song in the specific playlist is calculated by a method of averaging the actual playback lengths of the users of the electronic device for the songs corresponding to each song in the specific playlist in the acquired playlist. 13. The computer system according to claim 13, wherein the calculation is performed. The information management department
For each song in the playlist, additional information including at least one of the time, weather, and position at the time when the song is played is managed together with the playback history information.
The reproduction length prediction unit is
Using the song composition information, the playlist corresponding to the song composition of the specific playlist is acquired, and the playlist is obtained.
The expected playback length is calculated for each song of the specific playlist by using the playlist playback history information that matches the additional information at the time when the specific playlist is requested among the acquired playlists. 13. The computer system according to claim 13. The at least one processor
Further includes a data providing unit that distributes resources necessary for playing a song according to the playing length of each song of the specific playlist and provides sound source data for playing the specific playlist to the electronic device. , The computer system according to any one of claims 12 to 19.

Images