Intelligent music playing control system and method based on big data

Abstract

The invention provides an intelligent music playing control system and method based on big data, and relates to the technical field of big data analysis, and the method comprises the steps: carrying out the calculation according to user playing behavior data and scene data after data cleaning, and obtaining initial dynamic weight data; taking the initial dynamic weight data as a reference, adjusting the initial dynamic weight data according to the user behavior data variable quantity and the scene data variable quantity, and obtaining dynamic weight data based on adjustment; the method comprises the steps of obtaining user playing device types, identifying sound effect characteristic parameters of each device type, obtaining user sound effect preference parameters based on the obtained device types, and generating sound effect adaptation update data according to the sound effect characteristic parameters corresponding to different device types and different user sound effect preference parameters. According to the method, the playing behavior of the user in the recommended music is recorded, the behavior and situation change of the user are adapted in real time, and the matching of the playing sound effect and the playing device type is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of big data analysis, and in particular to a music intelligent playback control system and method based on big data. Background Art

[0002] With the popularity of music streaming platforms, personalized recommendation systems based on user data have gradually become mainstream. Existing music playback systems generate corresponding recommended playlists based on user data and play based on the specific sound effects selected by the user. Most of them ignore the differences between user devices and ignore the matching of user sound effect preferences. They provide the same sound effect experience without considering changes in user playback devices and situations, and are unable to adapt to users' playback sound effect needs. Summary of the Invention

[0003] The present invention provides a music intelligent playback control system and method based on big data, aiming to solve the technical problem of ignoring the distinguishing characteristics between user devices in the related art, and to solve the subsequent problem of difficulty in adapting to the user's playback sound effect needs.

[0004] In order to achieve the above objectives, the embodiments of the present application are implemented in the following manner: In a first aspect, an embodiment of the present application provides a method for intelligent music playback control based on big data, comprising the following steps: Obtain user playback behavior data and scenario data, and perform data cleaning on the user playback behavior data and scenario data; Calculate the initial dynamic weight data based on the user playback behavior data and scenario data after data cleaning; Taking the initial dynamic weight data as a benchmark, the initial dynamic weight data is adjusted according to the changes in user behavior data and scenario data, and dynamic weight data is obtained based on the adjustment; Obtain users' playback preference data at different time periods and generate a time-based weight model that adapts to the time series; Obtain the user's playback device type, identify the sound effect characteristic parameters of each device type, obtain the user's sound effect preference parameters based on the obtained device type, and generate sound effect adaptation update data based on the sound effect characteristic parameters corresponding to different device types and different user sound effect preference parameters; The final recommendation weight is generated by combining dynamic weight data, time-based weight model, and sound effect adaptation update data, and the music content with the highest priority is selected to generate recommended music; Record users' playback behavior of recommended music, and based on playback behavior feedback, optimize the recommended music and sound effect adaptation update data.

[0005] In conjunction with the first aspect, in some specific implementations, obtaining user playback behavior data and scenario data, and performing data cleaning on the user playback behavior data and scenario data specifically include: Obtain user playback behavior data and scenario data, clean the obtained user playback behavior data and scenario data, and obtain cleaned user playback behavior data and scenario data by removing invalid data, duplicate data, and abnormal data.

[0006] Furthermore, user playback behavior data includes playback frequency, dwell time, skipping behavior, likes behavior and repeated playback, and scenario data includes playback device type, time period, geographic location and user activity status.

[0007] In conjunction with the first aspect, in some specific implementations, calculations are performed based on the user playback behavior data and scenario data after data cleaning to obtain initial dynamic weight data, specifically including: The user's playback behavior data and scenario data are weighted and summed according to their respective weights, specifically: Where, Represents the weighted sum of user playback behavior data, Corresponding to the user's playback behavior data Dimensions The initial weight coefficient is used to control the influence of different playback behavior data when calculating the weight; Indicates the first dimensions, Indicates the number of different dimensions of user playback behavior data. represents the weighted sum of scenario data, Corresponding to the scenario data Dimensions The initial weight coefficient is used to control the influence of different scenario data when calculating the weight. Indicates the first dimensions, The number of different dimensions representing the scenario data.

[0008] Based on the weighted sum of the user's playback behavior data and scenario data according to their respective weights, the initial dynamic weight data is calculated using the following formula: Where, Indicates the initial dynamic weight data, which represents the increase or decrease of the user's behavior data of the corresponding dimension during the real-time update process. Represents the weighted sum of user playback behavior data, Corresponding to the user's playback behavior data Dimensions The initial weight coefficient is used to control the influence of different playback behavior data when calculating the weight; Indicates the first dimensions, Indicates the number of different dimensions of user playback behavior data. represents the weighted sum of scenario data, Corresponding to the scenario data Dimensions The initial weight coefficient is used to control the influence of different scenario data when calculating the weight. Indicates the first dimensions, The number of different dimensions representing the scenario data.

[0009] In conjunction with the first aspect, in some specific implementations, the initial dynamic weight data is used as a benchmark, and the initial dynamic weight data is adjusted according to the change in user behavior data and the change in scenario data. Based on the adjustment, dynamic weight data is obtained, and the adjustment formula is: Where, Indicates the dynamic weight data obtained after adjustment, Represents the initial dynamic weight data, Indicates the change in the corresponding scenario data The update coefficient is used to control the influence of the scenario data dimension during the real-time update process; Indicates the first The change in each dimension represents the increase or decrease in the scenario data of the corresponding dimension during the real-time update process of the user; Indicates the change in the corresponding user's playback behavior data The update coefficient is used to control the influence of the playback behavior data dimension during the real-time update process; Indicates the first dimensions, Indicates the number of different dimensions of user playback behavior data. Indicates the first The amount of change in a dimension.

[0010] Furthermore, based on the user's playback behavior data and scenario data, a weight is assigned to each data to generate initial dynamic weight data. When the user behavior and scenario change, the dynamic weight data is adjusted to reflect the user's immediate preferences in real time in the recommended content. The dynamic weight data adjusted in real time based on user behavior has high adaptability, can improve the flexibility of the recommendation system, and realize the accurate push of personalized recommended content.

[0011] In conjunction with the first aspect, in some specific implementations, the user's playback preference data for different time periods is obtained to generate a time-based weight model that adapts to the time series. The generation formula is: In the formula, by introducing different time periods Accumulate calculations to generate a time-based weight model that adapts to time series; Represents a time-based weight model that integrates user preference data from all time periods; Indicates the Dimensional data in the time period The weight coefficient in , Indicates that the user is in the time period The Playback preference data, Indicates the total number of time periods during which the user plays data. Indicates the number of behavioral data dimensions in each period, Represents the time period index, which is used to distinguish data from different time periods; Dimension index representing behavioral data.

[0012] Furthermore, the user's playback behavior data is recorded according to different time periods, and a time-based weight model for each time period is established. Through a weighted formula, the time-based weight model is combined with the dynamic weight data, so that the recommendation results have both time period characteristics and user's personalized preferences. The time-based weight model can help the system capture changes in user preferences in different time periods, and combine with dynamic weight data to generate dual-adaptive recommendation content, which helps to further improve user experience.

[0013] In conjunction with the first aspect, in some specific implementations, obtaining the user's playback device type, identifying the sound effect characteristic parameters of each device type, obtaining the user's sound effect preference parameters based on the obtained device type, and generating sound effect adaptation update data based on the sound effect characteristic parameters corresponding to different device types and different user sound effect preference parameters specifically include: Obtain the user's playback device type, which includes headphones, speakers, and car-mounted devices. Identify the sound effect characteristic parameters of each device type and obtain the user's sound effect preference parameters based on the obtained device type.

[0014] The initial sound effect adaptation data is generated based on the sound effect characteristic parameters corresponding to different device types and the sound effect preference parameters of different users. The generation formula is: Where, Indicates the impact of each device type's sound characteristics on initial sound adaptation. Indicates the impact of user sound effect preferences on initial sound effect adaptation, Indicates the device type sound effect characteristic parameters. Indicates the weight coefficient of the sound effect characteristic parameter setting for each device type. Indicates the user's sound effect preference parameters. Represents the weight coefficient of each user’s preference parameter setting, For the initial sound effect adaptation data, Indicates the number of dimensions of the device type sound effect characteristic parameters. The number of dimensions representing the user's sound effect preference parameters, Indicates the dimension index of the device type sound effect characteristic parameters. Dimension index representing the user's sound effect preference parameters.

[0015] Based on the dynamic weight data obtained after adjustment, the initial sound effect adaptation data is synchronously updated to generate sound effect adaptation update data. The update formula is: Indicates the sound effect adaptation update data obtained after synchronization update; Represents the initial sound effect adaptation data, which is used as the basis for the sound effect adaptation data; Indicates the update process The incremental weight coefficient of the sound effect characteristics of each device type is used to control the adjustment range of the sound effect characteristics on the sound effect adaptation data; Indicates the device type sound adaptation data dimensions; Indicates the update process The incremental weight coefficient of the user's sound effect adaptation data is used to control the adjustment range of the sound effect adaptation data by the preference; Indicates the user's sound effect preference parameters. dimensions, Indicates the number of dimensions of the device type sound effect adaptation data. The number of dimensions representing the user's sound effect preference parameters; Represents the adjustment item, which is used as the incremental part to synchronously update the sound effect adaptation data; Indicates the dimension index of the device type sound effect adaptation data. Dimension index representing the user's sound effect preference parameters.

[0016] Furthermore, the sound characteristics of the user's playback device are identified, and sound adaptation data is generated by combining the device characteristics and the user's sound preference parameters, and is adjusted synchronously with the dynamic weight data. By dynamically adjusting the sound adaptation data, the sound characteristics of the user's device can be adapted to ensure that the recommended music is highly consistent with the characteristics of the playback device, further enhancing the user's sound experience.

[0017] In conjunction with the first aspect, in some specific implementations, the dynamic weight data, the time-based weight model, and the sound effect adaptation update data are combined to generate the final recommendation weight, and the music content with the highest priority is screened to generate the recommended music, specifically including: The final recommendation weight is generated by combining the dynamic weight data, the time period weight model, and the weight values ​​of the sound effect adaptation update data. The generation formula is: Where, represents the final recommendation weight, Indicates the sound effect adaptation data that is updated synchronously with the dynamic weight data, that is, the sound effect adaptation update data. represents the periodized weight model, Represents dynamic weight data, 、 and These are weight adjustment coefficients, used to balance the impact of dynamic weight data, time period weight model, and sound effect adaptation data on the recommendation results.

[0018] Based on the final recommendation weight, the music content is filtered according to the weight priority, the music content with the highest priority is used as the recommended music, and the recommended music is sent to the user's playback device for playback.

[0019] Furthermore, by generating the final recommendation weight, filtering the music with the highest priority according to the final recommendation weight, and generating recommended music, we ensure the multi-dimensional optimization of the recommended content in terms of personalization, time period and sound effect adaptation, and significantly improve the user's music recommendation experience.

[0020] In conjunction with the first aspect, in some specific implementations, recording the user's playback behavior of recommended music, and optimizing the recommended music and sound effect adaptation update data based on the playback behavior feedback, specifically including: Convert the user's playback behavior in the recommended music into feedback data, assign corresponding feedback weights to different types of playback behaviors, and generate a set of feedback parameters that reflect user preferences .

[0021] According to the feedback parameter set , optimize the final recommendation weight: Where, is the final recommendation weight after optimization, represents the final recommendation weight, is the feedback weight coefficient, Indicates the feedback data of each playback behavior. Indicates the dimension index of the feedback data in the recommendation weight.

[0022] Optimize the sound effect adaptation and update data based on the feedback data of user playback behavior. The optimization formula is: Where, Indicates sound effect adaptation optimization data. Indicates that the sound effect adaptation update data, Optimize the coefficient for sound feedback, Indicates the degree of influence of each feedback data on the sound effect parameters. Indicates the dimension index of feedback data in sound effect adaptation. Indicates the total amount of user playback behavior feedback data.

[0023] Apply the optimized final recommendation weight to the user side, update the recommended music, and apply the sound effect adaptation optimization data to the user side playback settings to achieve adaptation of the playback sound effects to the playback device type.

[0024] Furthermore, through continuous optimization, the recommended music and sound effect parameters are continuously adjusted according to the user's immediate feedback, so as to achieve dynamic improvement of personalized recommendation effect and ensure the continuity of the user's playback experience.

[0025] In a second aspect, the present application further provides a music intelligent playback control system based on big data, comprising the following units: The data acquisition and conversion unit acquires user playback behavior data and scenario data, and performs data cleaning on the user playback behavior data and scenario data.

[0026] The calculation unit is used to calculate based on the user playback behavior data and scenario data after data cleaning to obtain initial dynamic weight data.

[0027] The dynamic weight unit is used to adjust the initial dynamic weight data based on the change in user behavior data and scenario data, and obtain dynamic weight data based on the adjustment.

[0028] The time period weight unit is used to obtain the user's playback preference data at different time periods and generate a time period weight model that adapts to the time series.

[0029] The sound effect adaptation unit is used to obtain the user's playback device type, identify the sound effect characteristic parameters of each device type, obtain the user's sound effect preference parameters based on the obtained device type, and generate sound effect adaptation update data according to the sound effect characteristic parameters corresponding to different device types and different user sound effect preference parameters.

[0030] The fusion output unit is used to combine dynamic weight data, periodic weight model and sound effect adaptation update data to generate the final recommendation weight, screen the music content with the highest priority and generate recommended music.

[0031] The optimization execution unit is used to record the user's playback behavior of recommended music, and based on the playback behavior feedback, optimize the recommended music and sound effect adaptation update data.

[0032] The beneficial effects of the technical solution of the present invention are: adjusting the recommended music through dynamic weight data, so that the recommended music can respond to changes in user needs in real time, integrating the user's preference behavior in different time periods with personalized dynamic weight data, and realizing dual adaptation of time period characteristics and user personalized preferences, generating sound effect adaptation data according to the user's playback device type and personalized sound effect preference parameters, and being able to better meet the user's device characteristics and preference needs in terms of sound effects, by recording the user's playback behavior in recommended music, adapting to the user's behavior and situational changes in real time, and realizing adaptation of playback sound effects and playback device types. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 A flowchart of a method for intelligent music playback control based on big data provided in an embodiment of the present application.

[0035] Figure 2 A unit framework diagram of a big data-based music intelligent playback control system provided in an embodiment of the present application.

[0036] Figure numerals: 20, data acquisition and conversion unit; 21, calculation unit; 22, dynamic weight unit; 23, time period weight unit; 24, sound effect adaptation unit; 25, fusion output unit; 26, optimization execution unit. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0038] See also Figure 1 , Figure 1 A flowchart of a method for intelligent music playback control based on big data is provided for an embodiment of the present application.

[0039] In this embodiment, a music intelligent playback control method based on big data may include step S100, step S200, step S300, step S400, step S500, step S600 and step S700.

[0040] Step S100: Obtain user playback behavior data and scenario data, and perform data cleaning on the user playback behavior data and scenario data.

[0041] Here, specifically: Obtain user playback behavior data and scenario data, clean the obtained user playback behavior data and scenario data, and obtain cleaned user playback behavior data and scenario data by removing invalid data, duplicate data, and abnormal data.

[0042] It should be noted that user playback behavior data includes playback frequency, dwell time, skipping behavior, likes behavior and repeated playback, and scenario data includes playback device type, time period, geographic location and user activity status.

[0043] It should be further explained that the data cleaning standards are dynamically adjusted according to user behavior patterns, those sudden behaviors are marked in abnormal data detection, and abnormal data that may affect the recommendation results are eliminated.

[0044] Step S200: Calculate based on the user play behavior data and scenario data after data cleaning to obtain initial dynamic weight data.

[0045] Here, specifically: The user's playback behavior data and scenario data are weighted and summed according to their respective weights, specifically: Where, Represents the weighted sum of user playback behavior data, Corresponding to the user's playback behavior data Dimensions The initial weight coefficient is used to control the influence of different playback behavior data when calculating the weight; Indicates the first dimensions, Indicates the number of different dimensions of user playback behavior data. represents the weighted sum of scenario data, Corresponding to the scenario data Dimensions The initial weight coefficient is used to control the influence of different scenario data when calculating the weight. Indicates the first dimensions, The number of different dimensions representing the scenario data.

[0046] Based on the weighted sum of the user's playback behavior data and scenario data according to their respective weights, the initial dynamic weight data is calculated using the following formula: Where, Indicates the initial dynamic weight data, which represents the increase or decrease of the user's behavior data of the corresponding dimension during the real-time update process. Represents the weighted sum of user playback behavior data, Corresponding to the user's playback behavior data Dimensions The initial weight coefficient is used to control the influence of different playback behavior data when calculating the weight; Indicates the first dimensions, Indicates the number of different dimensions of user playback behavior data. represents the weighted sum of scenario data, Corresponding to the scenario data Dimensions The initial weight coefficient is used to control the influence of different scenario data when calculating the weight. Indicates the first dimensions, The number of different dimensions representing the scenario data.

[0047] Step S300: Based on the initial dynamic weight data, the initial dynamic weight data is adjusted according to the change in user behavior data and the change in scenario data, and dynamic weight data is obtained based on the adjustment. The adjustment formula is: Where, Indicates the dynamic weight data obtained after adjustment, Represents the initial dynamic weight data, Indicates the change in the corresponding scenario data The update coefficient is used to control the influence of the scenario data dimension during the real-time update process; Indicates the first The change in each dimension represents the increase or decrease in the scenario data of the corresponding dimension during the real-time update process of the user; Indicates the change in the corresponding user's playback behavior data The update coefficient is used to control the influence of the playback behavior data dimension during the real-time update process; Indicates the first dimensions, Indicates the number of different dimensions of user playback behavior data. Indicates the first The amount of change in a dimension.

[0048] It should be noted that based on the user's playback behavior data and scenario data, a weight is assigned to each data item to generate initial dynamic weight data. When the user behavior and scenario change, the dynamic weight data is adjusted to reflect the user's immediate preferences in real time in the recommended content. The dynamic weight data adjusted in real time based on user behavior has high adaptability, can enhance the flexibility of the recommendation system, and realize the accurate push of personalized recommended content.

[0049] Step S400: Obtain the user's playback preference data for different time periods and generate a time-based weight model that adapts to the time series. The generation formula is: In the formula, by introducing different time periods Accumulate calculations to generate a time-based weight model that adapts to time series; Represents a time-based weight model that integrates user preference data from all time periods; Indicates the Dimensional data in the time period The weight coefficient in , Indicates that the user is in the time period The Playback preference data, Indicates the total number of time periods during which the user plays data. Indicates the number of behavioral data dimensions in each period, Represents the time period index, which is used to distinguish data from different time periods; Dimension index representing behavioral data.

[0050] It should be noted that the user's playback behavior data is recorded according to different time periods, and a time-based weight model for each time period is established. Through a weighted formula, the time-based weight model is combined with the dynamic weight data, so that the recommendation results have both time period characteristics and user's personalized preferences. The time-based weight model can help the system capture changes in user preferences in different time periods, and combine with dynamic weight data to generate dual-adapted recommendation content, which helps to further improve the user experience.

[0051] It should be further explained that when the playback device is switched, the sound effect adaptation data will be regenerated and the current sound effect parameters will be maintained to adapt to the user's sound effect preferences on different devices, thereby providing seamless switching.

[0052] Step S500: Obtain the user's playback device type, identify the sound effect characteristic parameters of each device type, obtain the user's sound effect preference parameters based on the obtained device type, and generate sound effect adaptation update data based on the sound effect characteristic parameters corresponding to different device types and different user sound effect preference parameters.

[0053] Here, specifically: Obtain the user's playback device type, which includes headphones, speakers, and car-mounted devices. Identify the sound effect characteristic parameters of each device type and obtain the user's sound effect preference parameters based on the obtained device type.

[0054] The initial sound effect adaptation data is generated based on the sound effect characteristic parameters corresponding to different device types and the sound effect preference parameters of different users. The generation formula is: Where, Indicates the impact of each device type's sound characteristics on initial sound adaptation. Indicates the impact of user sound effect preferences on initial sound effect adaptation, Indicates the device type sound effect characteristic parameters. Indicates the weight coefficient of the sound effect characteristic parameter setting for each device type. Indicates the user's sound effect preference parameters. Represents the weight coefficient of each user’s preference parameter setting, For the initial sound effect adaptation data, Indicates the number of dimensions of the device type sound effect characteristic parameters. The number of dimensions representing the user's sound effect preference parameters, Indicates the dimension index of the device type sound effect characteristic parameters. Dimension index representing the user's sound effect preference parameters.

[0055] Based on the dynamic weight data obtained after adjustment, the initial sound effect adaptation data is synchronously updated to generate sound effect adaptation update data. The update formula is: Indicates the sound effect adaptation update data obtained after synchronization update; Represents the initial sound effect adaptation data, which is used as the basis for the sound effect adaptation data; Indicates the update process The incremental weight coefficient of the sound effect characteristics of each device type is used to control the adjustment range of the sound effect characteristics on the sound effect adaptation data; Indicates the device type sound adaptation data dimensions; Indicates the update process The incremental weight coefficient of the user's sound effect adaptation data is used to control the adjustment range of the sound effect adaptation data by the preference; Indicates the user's sound effect preference parameters. dimensions, Indicates the number of dimensions of the device type sound effect adaptation data. The number of dimensions representing the user's sound effect preference parameters; Represents the adjustment item, which is used as the incremental part to synchronously update the sound effect adaptation data; Indicates the dimension index of the device type sound effect adaptation data. Dimension index representing the user's sound effect preference parameters.

[0056] It should be noted that by identifying the sound characteristics of the user's playback device, combining the device characteristics and the user's sound preference parameters to generate sound adaptation data, and adjusting it synchronously with the dynamic weight data, by dynamically adjusting the sound adaptation data, it is possible to adapt to the user's device sound characteristics, ensure that the recommended music is highly consistent with the playback device characteristics, and further enhance the user's sound experience.

[0057] It should be further explained that after the recommended music is generated, it will be continuously updated to ensure that the recommendation results remain fresh during long-term playback, avoiding a decline in user experience due to uniformity.

[0058] Step S600: Combine the dynamic weight data, the time-based weight model, and the sound effect adaptation update data to generate the final recommendation weight, and filter the music content with the highest priority to generate recommended music.

[0059] Here, specifically: The final recommendation weight is generated by combining the dynamic weight data, the time period weight model, and the weight values ​​of the sound effect adaptation update data. The generation formula is: Where, represents the final recommendation weight, Indicates the sound effect adaptation data that is updated synchronously with the dynamic weight data, that is, the sound effect adaptation update data. represents the periodized weight model, Represents dynamic weight data, 、 and These are weight adjustment coefficients, used to balance the impact of dynamic weight data, time period weight model, and sound effect adaptation data on the recommendation results.

[0060] Based on the final recommendation weight, the music content is filtered according to the weight priority, the music content with the highest priority is used as the recommended music, and the recommended music is sent to the user's playback device for playback.

[0061] It should be noted that by generating the final recommendation weight, filtering the music with the highest priority based on the final recommendation weight, and generating recommended music, we ensure multi-dimensional optimization of the recommended content in terms of personalization, time period, and sound effect adaptation, thereby significantly improving the user's music recommendation experience.

[0062] It should be further explained that the dynamic weight data and sound effect adaptation data will be updated based on feedback data so that the recommended music and sound effect parameters can continue to meet user needs, ensuring the accuracy and diversity of the recommendation results under different usage intensities and preference stability.

[0063] Step S700: Record the user's playback behavior of the recommended music, and based on the playback behavior feedback, optimize the recommended music and sound effect adaptation update data.

[0064] Here, specifically: Convert the user's playback behavior in the recommended music into feedback data, assign corresponding feedback weights to different types of playback behaviors, and generate a set of feedback parameters that reflect user preferences .

[0065] According to the feedback parameter set , optimize the final recommendation weight: Where, is the final recommendation weight after optimization, represents the final recommendation weight, is the feedback weight coefficient, Indicates the feedback data of each playback behavior. Indicates the dimension index of the feedback data in the recommendation weight.

[0066] Optimize the sound effect adaptation and update data based on the feedback data of user playback behavior. The optimization formula is: Where, Indicates sound effect adaptation optimization data. Indicates that the sound effect adaptation update data, Optimize the coefficient for sound feedback, Indicates the degree of influence of each feedback data on the sound effect parameters. Indicates the dimension index of feedback data in sound effect adaptation. Indicates the total amount of user playback behavior feedback data.

[0067] Apply the optimized final recommendation weight to the user side, update the recommended music, and apply the sound effect adaptation optimization data to the user side playback settings to achieve adaptation of the playback sound effects to the playback device type.

[0068] It should be noted that through continuous optimization, the recommended music and sound effect parameters are continuously adjusted according to the user's immediate feedback, so as to achieve dynamic improvement of personalized recommendation effect and ensure the continuity of the user's playback experience.

[0069] At this point, through the above steps, a music intelligent playback control method based on big data is completed.

[0070] In addition, please refer to Figure 2 , Figure 2 A unit framework diagram of a big data-based music intelligent playback control system provided in an embodiment of the present application.

[0071] This application also provides a music intelligent playback control system based on big data, including the following units: The data acquisition and conversion unit 20 is used to acquire multidimensional data, clean and label the acquired multidimensional data, and use the labeled multidimensional data as input for the dynamic weight curve and the periodized weight model.

[0072] The calculation unit 21 is used to perform calculations based on the user's play behavior data and scenario data after data cleaning to obtain initial dynamic weight data.

[0073] The dynamic weight unit 22 is used to adjust the initial dynamic weight data based on the change amount of the user behavior data and the change amount of the scenario data based on the initial dynamic weight data, and obtain dynamic weight data based on the adjustment.

[0074] The time period weighting unit 23 is used to obtain the user's playback preference data at different time periods and generate a time period weighting model that adapts to the time series.

[0075] The sound effect adaptation unit 24 is used to obtain the user's playback device type, identify the sound effect characteristic parameters of each device type, obtain the user's sound effect preference parameters based on the obtained device type, and generate sound effect adaptation update data according to the sound effect characteristic parameters corresponding to different device types and different user sound effect preference parameters.

[0076] The fusion output unit 25 is used to combine the dynamic weight data, the time period weight model and the sound effect adaptation update data to generate the final recommendation weight, and to screen the music content with the highest priority to generate recommended music.

[0077] The optimization execution unit 26 is used to record the user's playback behavior of the recommended music, and based on the playback behavior feedback, optimize the recommended music and sound effect adaptation update data.

[0078] When the system is actually put into use, the user's playback behavior data and scenario data are first obtained through the data acquisition and conversion unit 20. The user's playback behavior data includes playback frequency, dwell time, skipping behavior, like behavior and repeated playback, while the scenario data includes the user's playback device type, time period, geographic location and user activity status. After obtaining the data, the system performs a cleaning operation to remove invalid, duplicate and abnormal data.

[0079] The system uses the calculation unit 21 and the dynamic weight unit 22 to assign weights to different types of behavior and scenario data based on the cleaned data, generates initial dynamic weight data, and personalizes the recommended music based on the dynamic weight data. During operation, the dynamic weight data will be adjusted in real time according to changes in the user's playback behavior. When a user frequently likes or skips a certain type of music in a short period of time, the system will increase or decrease the weight of this type of music accordingly to ensure that the recommendation results meet the user's immediate preferences.

[0080] Based on the time period weight unit 23, the system generates a corresponding time period weight model by recording the user's preference data in different time periods to ensure that the recommended content can adapt to the user's time preference. The system records the user's playback preferences in different time periods and generates a corresponding time period weight model; the system integrates the time period weight model with the dynamic weight data to form a weight structure suitable for different time periods, so that the recommended content conforms to the user's preferences in a specific time period.

[0081] According to the sound effect adaptation unit 24, the system generates sound effect adaptation data for the user's playback device type and personalized sound effect preference parameters. The system detects the user's playback device type and identifies the device's sound effect characteristics, which include bass enhancement, stereo effect, and dynamic range. The system combines the device's sound effect characteristics with the user's sound effect preference parameters to generate initial sound effect adaptation data, and synchronizes the dynamic weight curve in real time during playback for adjustment.

[0082] According to the fusion output unit 25, the system will combine the dynamic weight data, the time period weight model and the sound effect adaptation data to generate the final recommendation weight; based on the final recommendation weight, the music with the highest priority is screened out, the recommended music is generated and sent to the user end.

[0083] According to the optimization execution unit 26, the system applies the optimized recommendation weight to the user end, updates the recommended music, and applies the adjusted sound effect parameters to the user end playback settings to achieve adaptation of the playback sound effect to the playback device type.

[0084] At this point, through the above units, a music intelligent playback control system based on big data has been completed.

[0085] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A music intelligent playback control method based on big data, characterized in that: The steps include: Obtain user playback behavior data and scenario data, and perform data cleaning on the user playback behavior data and scenario data; Calculate the initial dynamic weight data based on the user playback behavior data and scenario data after data cleaning; Taking the initial dynamic weight data as a benchmark, the initial dynamic weight data is adjusted according to the changes in user behavior data and scenario data, and dynamic weight data is obtained based on the adjustment; Obtain users' playback preference data at different time periods and generate a time-based weight model that adapts to the time series; Obtain the user's playback device type, identify the sound effect characteristic parameters of each device type, obtain the user's sound effect preference parameters based on the obtained device type, and generate sound effect adaptation update data based on the sound effect characteristic parameters corresponding to different device types and different user sound effect preference parameters; The final recommendation weight is generated by combining dynamic weight data, time-based weight model, and sound effect adaptation update data, and the music content with the highest priority is selected to generate recommended music; Record users' playback behavior of recommended music, and based on playback behavior feedback, optimize the recommended music and sound effect adaptation update data.

2. The method for intelligent music playback control based on big data according to claim 1, characterized in that: Obtain user playback behavior data and scenario data, and perform data cleaning on the user playback behavior data and scenario data, specifically including: Obtain user playback behavior data and scenario data, clean the obtained user playback behavior data and scenario data, and obtain cleaned user playback behavior data and scenario data by removing invalid data, duplicate data, and abnormal data.

3. The method for intelligent music playback control based on big data according to claim 1, characterized in that: Calculate the initial dynamic weight data based on the user playback behavior data and scenario data after data cleaning, including: The user's playback behavior data and scenario data are weighted and summed according to their respective weights, specifically: ； ； Where, Represents the weighted sum of user playback behavior data, Corresponding to the user's playback behavior data Dimensions The initial weight coefficient is used to control the influence of different playback behavior data when calculating the weight; Indicates the first dimensions, Indicates the number of different dimensions of user playback behavior data. represents the weighted sum of scenario data, Corresponding to the scenario data Dimensions The initial weight coefficient is used to control the influence of different scenario data when calculating the weight. Indicates the first dimensions, The number of different dimensions representing the scenario data; Based on the weighted sum of the user's playback behavior data and scenario data according to their respective weights, the initial dynamic weight data is calculated using the following formula: ； Where, Indicates the initial dynamic weight data, which represents the increase or decrease of the user's behavior data of the corresponding dimension during the real-time update process. Represents the weighted sum of user playback behavior data, Corresponding to the user's playback behavior data Dimensions The initial weight coefficient is used to control the influence of different playback behavior data when calculating the weight; Indicates the first dimensions, Indicates the number of different dimensions of user playback behavior data. represents the weighted sum of scenario data, Corresponding to the scenario data Dimensions The initial weight coefficient is used to control the influence of different scenario data when calculating the weight. Indicates the first dimensions, The number of different dimensions representing the scenario data.

4. The method for intelligent music playback control based on big data according to claim 1, characterized in that: Taking the initial dynamic weight data as the benchmark, the initial dynamic weight data is adjusted according to the changes in user behavior data and scenario data. The dynamic weight data is obtained based on the adjustment. The adjustment formula is: ； Where, Indicates the dynamic weight data obtained after adjustment, Represents the initial dynamic weight data, Indicates the change in the corresponding scenario data The update coefficient is used to control the influence of the scenario data dimension during the real-time update process; Indicates the first The change in each dimension represents the increase or decrease in the scenario data of the corresponding dimension during the real-time update process of the user; Indicates the change in the corresponding user's playback behavior data The update coefficient is used to control the influence of the playback behavior data dimension during the real-time update process; Indicates the first dimensions, Indicates the number of different dimensions of user playback behavior data. Indicates the first The amount of change in a dimension.

5. The method for intelligent music playback control based on big data according to claim 1, characterized in that: Obtain user playback preference data at different time periods and generate a time-based weight model that adapts to the time series, specifically including: Obtain user playback preference data at different time periods and introduce different time periods Accumulate the calculation to generate a time-based weight model that adapts to the time series. The calculation formula is: ； Where, Represents a time-based weight model that integrates user preference data from all time periods; Indicates the Dimensional data in the time period The weight coefficient in , Indicates that the user is in the time period The Playback preference data, Indicates the total number of time periods during which the user plays data. Indicates the number of behavioral data dimensions in each period, Represents the time period index, which is used to distinguish data from different time periods; Dimension index representing behavioral data.

6. The method for intelligent music playback control based on big data according to claim 1, characterized in that: Obtain the user's playback device type, identify the sound effect characteristic parameters of each device type, obtain the user's sound effect preference parameters based on the obtained device type, and generate sound effect adaptation update data based on the sound effect characteristic parameters corresponding to different device types and different user sound effect preference parameters. Specifically, it includes: Obtain the user's playback device type, which includes headphones, speakers, and in-car devices. Identify the sound effect characteristic parameters of each device type and obtain the user's sound effect preference parameters based on the obtained device type. The initial sound effect adaptation data is generated based on the sound effect characteristic parameters corresponding to different device types and the sound effect preference parameters of different users. The generation formula is: ； Where, Indicates the impact of each device type's sound characteristics on initial sound adaptation. Indicates the impact of user sound effect preferences on initial sound effect adaptation, Indicates the device type sound effect characteristic parameters. Indicates the weight coefficient of the sound effect characteristic parameter setting for each device type. Indicates the user's sound effect preference parameters. Represents the weight coefficient of each user’s preference parameter setting, For the initial sound effect adaptation data, Indicates the number of dimensions of the device type sound effect characteristic parameters. The number of dimensions representing the user's sound effect preference parameters, Indicates the dimension index of the device type sound effect characteristic parameters. Dimension index representing the user's sound effect preference parameters; Based on the dynamic weight data obtained after adjustment, the initial sound effect adaptation data is synchronously updated to generate sound effect adaptation update data. The update formula is: ； Indicates the sound effect adaptation update data obtained after synchronization update; Represents the initial sound effect adaptation data, which is used as the basis for the sound effect adaptation data; Indicates the update process The incremental weight coefficient of the sound effect characteristics of each device type is used to control the adjustment range of the sound effect characteristics on the sound effect adaptation data; Indicates the device type sound adaptation data dimensions; Indicates the update process The incremental weight coefficient of the user's sound effect adaptation data is used to control the adjustment range of the sound effect adaptation data by the preference; Indicates the user's sound effect preference parameters. dimensions, Indicates the number of dimensions of the device type sound effect adaptation data. The number of dimensions representing the user's sound effect preference parameters; Indicates the dimension index of the device type sound effect adaptation data. Dimension index representing the user's sound effect preference parameters.

7. The method for intelligent music playback control based on big data according to claim 1, characterized in that: The final recommendation weight is generated by combining dynamic weight data, time-based weight model, and sound effect adaptation update data. The music content with the highest priority is selected to generate recommended music. Specifically, it includes: The final recommendation weight is generated by combining the dynamic weight data, the time period weight model, and the weight values ​​of the sound effect adaptation update data. The generation formula is: ； Where, represents the final recommendation weight, Indicates the sound effect adaptation data that is updated synchronously with the dynamic weight data, that is, the sound effect adaptation update data. represents the periodized weight model, Represents dynamic weight data, 、 and These are weight adjustment coefficients, used to balance the impact of dynamic weight data, time period weight model, and sound effect adaptation data on the recommendation results. Based on the final recommendation weight, the music content is filtered according to the weight priority, the music content with the highest priority is used as the recommended music, and the recommended music is sent to the user's playback device for playback.

8. The method for intelligent music playback control based on big data according to claim 1, characterized in that: Record users' playback behavior of recommended music, and based on playback behavior feedback, optimize the recommended music and sound effects adaptation update data, including: Convert the user's playback behavior in the recommended music into feedback data, assign corresponding feedback weights to different types of playback behaviors, and generate a set of feedback parameters that reflect user preferences ; According to the feedback parameter set , optimize the final recommendation weight: ； Where, is the final recommendation weight after optimization, represents the final recommendation weight, is the feedback weight coefficient, Indicates the feedback data of each playback behavior. Represents the dimension index of feedback data in recommendation weight; Optimize the sound effect adaptation and update data based on the feedback data of user playback behavior. The optimization formula is: ； Where, Indicates sound effect adaptation optimization data. Indicates that the sound effect adaptation update data, Optimize the coefficient for sound feedback, Indicates the degree of influence of each feedback data on the sound effect parameters. Indicates the dimension index of feedback data in sound effect adaptation. Indicates the total amount of user playback behavior feedback data; Apply the optimized final recommendation weight to the user side, update the recommended music, and apply the sound effect adaptation optimization data to the user side playback settings to achieve adaptation of the playback sound effects to the playback device type.

9. A music intelligent playback control system based on big data, characterized in that: Includes the following units: A data acquisition and conversion unit, which acquires user playback behavior data and scenario data, and performs data cleaning on the user playback behavior data and scenario data; A calculation unit, configured to calculate the initial dynamic weight data based on the user playback behavior data and scenario data after data cleaning; A dynamic weight unit, configured to adjust the initial dynamic weight data based on the change in user behavior data and the change in scenario data, taking the initial dynamic weight data as a benchmark, and obtain dynamic weight data based on the adjustment; The time period weight unit is used to obtain the user's playback preference data at different time periods and generate a time period weight model that adapts to the time series; A sound effect adaptation unit is used to obtain the user's playback device type, identify the sound effect characteristic parameters of each device type, obtain the user's sound effect preference parameters based on the obtained device type, and generate sound effect adaptation update data based on the sound effect characteristic parameters corresponding to different device types and different user sound effect preference parameters; The fusion output unit is used to combine dynamic weight data, time-based weight model, and sound effect adaptation update data to generate the final recommendation weight, screening the music content with the highest priority to generate recommended music; The optimization execution unit is used to record the user's playback behavior of recommended music, and based on the playback behavior feedback, optimize the recommended music and sound effect adaptation update data.

Images