Music data intelligent clustering system and method based on audio feature analysis

Abstract

The invention provides a music data intelligent clustering system and method based on audio feature analysis. The method comprises the following steps: acquiring updated music audio data of music software, and identifying beat period features and melody emotion features of the updated music audio data, identifying music categories corresponding to the updated music audio data according to the beat period features and the melody emotion features, and constructing clustering adaptation parameters according to category features corresponding to each music category, and determining the clustering similarity between the updated music audio data and each music category according to the clustering adaptation parameter, dividing the updated music audio data into the corresponding target music category, and classifying the music by combining a clustering technology with a feature recognition technology, thereby solving the defects of the traditional technology, and improving the music classification efficiency. And the music data classification accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and particularly to an intelligent clustering system and method for music data based on audio feature analysis. Background Art

[0002] With the development of multimedia and Internet information technologies, people's demand for retrieving audio information resources is getting stronger and stronger, which also makes the efficient classification of music information become a current research hotspot. At the same time, with the development of the Internet, the amount of music data has exploded. There are 35 million different songs on average in each music platform, and a large number of albums are continuously released on the Internet every week. This makes it difficult for traditional statistical and computational methods to quickly classify a large amount of music.

[0003] Currently, although clustering analysis is applied in many fields, there are many deficiencies in existing music data clustering methods, such as inaccurate feature extraction and poor clustering algorithm effects, which cannot meet the requirements of music recommendation and classification management and cannot respond to the rapidly developing Internet.

[0004] Therefore, the present invention provides an intelligent clustering system and method for music data based on audio feature analysis. Summary of the Invention

[0005] The intelligent clustering system and method for music data based on audio feature analysis of the present invention uses clustering technology combined with feature recognition technology to classify music, solves the drawbacks of traditional technologies, and improves the accuracy of music data classification.

[0006] The present invention provides an intelligent clustering system for music data based on audio feature analysis, including:

[0007] A data acquisition module for acquiring updated music audio data of a music software;

[0008] A feature extraction module for identifying the beat period feature and the melody emotion feature of the updated music audio data;

[0009] A clustering processing module for identifying the music category corresponding to the updated music audio data according to the beat period feature and the melody emotion feature, and constructing a clustering adaptation parameter according to the category feature corresponding to each music category;

[0010] A clustering optimization module for determining the clustering similarity between the updated music audio data and each music category according to the clustering adaptation parameter, and dividing the updated music audio data into the corresponding target music category.

[0011] In an implementable manner,

[0012] The data acquisition module includes:

[0013] A data capture unit for capturing real-time complete update data of the music software;

[0014] A data screening unit for performing duration analysis and type analysis on each piece of the real-time complete update data respectively, and screening the updated music audio data that meets the specified duration and specified type.

[0015] In an implementable manner,

[0016] The feature extraction module includes:

[0017] A data processing unit for performing Fourier transform on the updated music audio data to obtain a number of frequency components, counting the frequency times corresponding to each frequency component, determining the main frequency component of the updated music audio data, and respectively plotting each frequency component in the frequency domain space to obtain the phase information corresponding to each frequency component;

[0018] A beat determination unit for determining a number of single beat features of the updated music audio data based on the frequency difference between each frequency component and the main frequency component, determining the main beat feature of the updated music audio data based on the main frequency component, and using each single beat feature to correct the main beat feature to obtain the beat information of the updated music audio data;

[0019] A period determination unit for generating an initial period according to the main phase information corresponding to the main frequency component, decomposing the beat information by using the initial period to obtain a number of decomposed beats, and adjusting the period range of the initial period according to the beat differences between different decomposed beats;

[0020] A melody determination unit for decomposing the updated music audio data to obtain the melody features corresponding to the updated music audio data in different melody dimensions, and determining the core melody of the updated music audio data according to the feature values corresponding to each melody feature;

[0021] An emotion determination unit for performing text recognition on the updated music audio data by using natural language processing technology to obtain the corresponding data text, performing emotion annotation on the data text by using a preset emotion dictionary to determine the emotion tendency of the updated music audio data, and training the emotion tendency by using the random forest method to obtain the data emotion of the updated music audio data;

[0022] A feature determination unit for generating the beat period feature of the updated music audio data according to the beat information and the period range, and generating the melody emotion feature of the updated music audio data according to the core melody and the data emotion.

[0023] In an implementable manner,

[0024] The melody determination unit includes:

[0025] A melody decomposition subunit, configured to decompose the updated music audio data to obtain pitch characteristics, volume characteristics, speech rate characteristics, duration characteristics, and stress characteristics of the updated music audio data;

[0026] A numerical analysis subunit, configured to analyze the numerical levels of the corresponding characteristic values according to the numerical grading criteria corresponding to each melody dimension, and combine the numerical levels to generate combined melody information of the updated music audio data;

[0027] A melody reconstruction subunit, configured to construct a pure audio according to the combined melody information, and generate a core melody of the updated music audio data according to the pure melody of the pure audio.

[0028] In an implementable manner,

[0029] The clustering processing module includes:

[0030] A preliminary classification unit, configured to match a corresponding first music category set for the updated music audio data according to the beat period characteristic, match a corresponding second music category set for the updated music audio data according to the melody emotion characteristic, and determine the corresponding initial music category of the updated music audio data according to the intersection between the first music category set and the second music category set;

[0031] A clustering analysis unit, configured to respectively obtain the existing music audio data corresponding to each initial music category, construct a category attribute graph corresponding to each initial music category based on the music labels corresponding to each existing music audio data, input the updated music audio data into the category attribute graph for classification verification, obtain the matching nodes of each updated music audio data in each category attribute graph, and screen out the music categories whose node distances between the matching nodes and the corresponding existing nodes are less than a specified distance;

[0032] A parameter generation unit, configured to generate category characteristics corresponding to the music categories according to the category attribute graph, perform fitting matching between the beat period characteristic and the category characteristics to generate beat period parameters of the updated music audio data in the corresponding music categories, perform fitting matching between the melody emotion characteristic and the category characteristics to generate melody emotion parameters of the updated music audio data in the corresponding music categories, and perform data fusion on the beat period parameters and the melody emotion parameters to generate clustering adaptation parameters of the updated music audio data in each music category.

[0033] In an implementable manner,

[0034] The preliminary classification unit is further configured to:

[0035] When there is no intersection between the first music category set and the second music category set, determine that the updated music audio data is abnormal, and generate a re-acquisition instruction;

[0036] Use the re-acquisition instruction to control the data acquisition module to re-acquire the updated music audio data.

[0037] In an implementable manner,

[0038] The clustering optimization module includes:

[0039] A clustering verification unit, configured to iteratively cluster the updated music audio data with the corresponding music category based on the clustering adaptation parameter to obtain the data similarity features between the updated music audio data and the existing music audio data in each music category;

[0040] A clustering optimization unit, configured to count a plurality of data similarity features corresponding to each music category to determine the clustering similarity between the updated music audio data and each music category, and determine a plurality of target music categories corresponding to each updated music audio data based on the clustering similarity;

[0041] A division execution unit, configured to respectively screen relevant music audio data in each target music category whose clustering similarity with the updated music audio data is higher than the standard similarity, set a recommendation guide for the updated music audio data, and divide the updated music audio data into the corresponding target music categories respectively.

[0042] In an implementable manner,

[0043] It further includes:

[0044] After updating the updated music audio data to the corresponding target music category, regard the updated music audio data as existing music audio data.

[0045] The present invention provides an intelligent clustering method for music data based on audio feature analysis, including:

[0046] Step 1: Collect updated music audio data of a music software;

[0047] Step 2: Identify the beat period feature and the melody emotion feature of the updated music audio data;

[0048] Step 3: Identify the music category corresponding to the updated music audio data according to the beat period feature and the melody emotion feature, and construct a clustering adaptation parameter according to the category feature corresponding to each music category;

[0049] Step 4: Determine the clustering similarity between the updated music audio data and each music category according to the clustering adaptation parameter, and divide the updated music audio data into the corresponding target music category.

[0050] In an implementable manner,

[0051] The said step 2 includes:

[0052] Step 21: Perform Fourier transform on the updated music audio data to obtain a number of frequency components, count the frequency times corresponding to each frequency component, determine the main frequency component of the updated music audio data, and respectively draw each frequency component in the frequency domain space to obtain the phase information corresponding to each frequency component;

[0053] Step 22: Determine a number of single beat features of the updated music audio data based on the frequency difference between each frequency component and the main frequency component, determine the main beat feature of the updated music audio data based on the main frequency component, and use each single beat feature to correct the main beat feature to obtain the beat information of the updated music audio data;

[0054] Step 23: Generate an initial period according to the main phase information corresponding to the main frequency component, decompose the beat information by using the initial period to obtain a number of decomposed beats, and adjust the period range of the initial period according to the beat difference between different decomposed beats;

[0055] Step 24: Decompose the updated music audio data into melody features corresponding to the updated music audio data in different melody dimensions, and determine the core melody of the updated music audio data according to the characteristic value corresponding to each melody feature;

[0056] Step 25: Use natural language processing technology to perform text recognition on the updated music audio data to obtain the corresponding data text, perform emotion annotation on the data text by using a preset emotion dictionary, determine the emotion tendency of the updated music audio data, and use the random forest method to train the emotion tendency to obtain the data emotion of the updated music audio data;

[0057] Step 26: Generate the beat period feature of the updated music audio data according to the beat information and the period range, and generate the melody emotion feature of the updated music audio data according to the core melody and the data emotion.

[0058] The achievable beneficial effects of the above technical solution are as follows: When the music audio data is updated in the music software, the data is collected in a timely manner and the beat period characteristics and melody emotion characteristics of the data are identified. Then, according to the beat period characteristics and melody emotion characteristics, the corresponding classification is matched for the updated music audio data, and the corresponding clustering adaptation parameters are constructed. Further, the parameters are used to optimize the clustering of the updated music audio data, and the target music category that the updated music audio data conforms to is determined. In this way, the updated music audio data in the music software can be captured in a timely manner, and then its necessary characteristics are identified, and then clustering is performed. In order to improve the accuracy of clustering, the clustering adaptation parameters between the updated music audio data and different music categories are determined, and then the parameters are used to deeply analyze the clustering similarity between the two, so as to determine the target music category of the updated music audio data, improve the accuracy of the traditional clustering technology, and can assign a suitable category to the updated music audio data, which is convenient for listeners to select their favorite music category and improve the listeners' love for the music software.

[0059] Other features and advantages of the present invention will be described in the following specification, and, in part, will be obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention can be achieved and obtained by the structure specifically pointed out in the written specification and the drawings.

[0060] The technical solution of the present invention will be further described in detail below with reference to the drawings and embodiments. Brief Description of the Drawings

[0061] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention. In the drawings:

[0062] Figure 1 It is a schematic diagram of the composition of the music data intelligent clustering system based on audio feature analysis in the embodiment of the present invention;

[0063] Figure 2 It is a schematic diagram of the working process of the music data intelligent clustering method based on audio feature analysis in the embodiment of the present invention. Detailed Embodiments

[0064] The following describes the preferred embodiments of the present invention with reference to the drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention and are not used to limit the present invention.

[0065] Embodiment 1

[0066] This embodiment provides a music data intelligent clustering system based on audio feature analysis, asFigure 1 As shown in

[0067] a data acquisition module, configured to acquire updated music audio data of a music software;

[0068] a feature extraction module, configured to identify the beat period feature and the melody emotion feature of the updated music audio data;

[0069] a clustering processing module, configured to identify the music category corresponding to the updated music audio data according to the beat period feature and the melody emotion feature, and construct a clustering adaptation parameter according to the category feature corresponding to each music category;

[0070] a clustering optimization module, configured to determine the clustering similarity between the updated music audio data and each music category according to the clustering adaptation parameter, and divide the updated music audio data into the corresponding target music category.

[0071] In this example, the beat period feature represents the rhythm basis and the repetition period of the updated music audio data;

[0072] In this example, the melody emotion feature represents the melody composed of the notes of the updated music audio data and the emotion expressed by it;

[0073] In this example, the clustering adaptation parameter represents the parameter generated when the updated music audio data is clustered into a music category and adapts to this music category;

[0074] In this example, the clustering similarity represents the similarity between the updated music audio data and the existing music audio data in the music category;

[0075] In this example, the number of target music categories can be one or more.

[0076] Working principle and beneficial effects of the above technical solution: When music audio data is updated in a music software, the data is collected in a timely manner, and the beat period characteristics and melody emotion characteristics of the data are identified. Then, corresponding classifications are matched for the updated music audio data according to the beat period characteristics and melody emotion characteristics, and corresponding clustering adaptation parameters are constructed. Further, the parameters are used to optimize the clustering of the updated music audio data, and the target music category that the updated music audio data conforms to is determined. In this way, the updated music audio data in the music software can be captured in a timely manner, and then its necessary characteristics are identified, and then clustering is performed. In order to improve the accuracy of clustering, the clustering adaptation parameters between the updated music audio data and different music categories are determined, and then the parameters are used to deeply analyze the clustering similarity between the two, so as to determine the target music category of the updated music audio data, improve the accuracy of traditional clustering technology, and can assign suitable categories to the updated music audio data, facilitating listeners to select their favorite music categories and improving the listeners' love for the music software.

[0077] Embodiment 2

[0078] Based on the Embodiment 1, for the intelligent music data clustering system based on audio feature analysis, the data acquisition module includes:

[0079] A data capture unit for capturing the real-time complete update data of the music software;

[0080] A data screening unit for performing duration analysis and type analysis on each piece of the real-time complete update data respectively, and screening the updated music audio data that meets the specified duration and specified type.

[0081] In this example, the real-time complete update data represents the complete data that has been updated;

[0082] In this example, the duration analysis means analyzing the music duration of the real-time complete update data, and the type analysis means identifying the data type of the real-time complete update data;

[0083] In this example, the specified duration is 20 seconds;

[0084] In this example, the specified type represents one or two of the music type and the audio type.

[0085] Working principle and beneficial effects of the above technical solution: Capture the updated data in the music software in a timely manner, then analyze it, and screen out the updated music audio data that meets the specified duration and specified type to wait for the next clustering and classification work. In this way, the real-time complete update data can be preliminarily screened to avoid irrelevant data from mixing in and interfering with the clustering quality and efficiency.

[0086] Embodiment 3

[0087] Based on Example 1, for the intelligent music data clustering system based on audio feature analysis, the feature extraction module includes:

[0088] A data processing unit, configured to perform Fourier transform on the updated music audio data to obtain a plurality of frequency components, count the frequency times corresponding to each frequency component, determine the main frequency component of the updated music audio data, and respectively draw each frequency component in the frequency domain space to obtain the phase information corresponding to each frequency component;

[0089] A beat determination unit, configured to determine a plurality of single beat features of the updated music audio data based on the frequency difference between each frequency component and the main frequency component, determine the main beat feature of the updated music audio data based on the main frequency component, and use each single beat feature to correct the main beat feature to obtain the beat information of the updated music audio data;

[0090] A period determination unit, configured to generate an initial period according to the main phase information corresponding to the main frequency component, decompose the beat information by using the initial period to obtain a plurality of decomposed beats, and adjust the period range of the initial period according to the beat difference between different decomposed beats;

[0091] A melody determination unit, configured to decompose the updated music audio data to obtain the melody features corresponding to the updated music audio data in different melody dimensions, and determine the core melody of the updated music audio data according to the feature values corresponding to each melody feature;

[0092] An emotion determination unit, configured to perform text recognition on the updated music audio data by using natural language processing technology to obtain corresponding data text, perform emotion annotation on the data text by using a preset emotion dictionary to determine the emotion tendency of the updated music audio data, and perform training on the emotion tendency by using the random forest method to obtain the data emotion of the updated music audio data;

[0093] A feature determination unit, configured to generate a beat period feature of the updated music audio data according to the beat information and the period range, and generate a melody emotion feature of the updated music audio data according to the core melody and the data emotion.

[0094] In this example, the frequency component represents the component about different frequencies included in the updated music audio data;

[0095] In this example, the frequency times represents the number of times a frequency component appears;

[0096] In this example, the main frequency component represents the frequency component that appears most frequently in the updated music audio data;

[0097] In this example, the phase information represents the waveform presented by the frequency component;

[0098] In this example, the frequency difference represents the phase shift between the frequency component and the main frequency component;

[0099] In this example, the single beat feature represents the beat presented by a frequency component;

[0100] In this example, the main beat feature represents the beat presented by the main frequency component;

[0101] In this example, the beat information represents the continuous beats presented by the updated music audio data;

[0102] In this example, the process of correcting the main beat feature is: the process of correcting the notes of the main beat feature using the single beat feature;

[0103] In this example, the initial period represents the period derived based on the main phase information;

[0104] In this example, the period range represents the range of the period length of the initial period;

[0105] In this example, the melody dimension includes the pitch dimension, volume dimension, speech rate dimension, duration dimension, and stress dimension;

[0106] In this example, the core melody represents the melody presented multiple times in the updated music audio data;

[0107] In this example, the emotion annotation represents the process of matching corresponding emotions to the data text;

[0108] In this example, the preset emotion dictionary contains several words, and each word is set with an emotion value. The emotion tendency of the text is identified by counting the emotion values contained in a text;

[0109] In this example, the data emotion represents the emotion presented by the updated music audio data.

[0110] Working principle and beneficial effects of the above technical solution: In order to be able to identify various features of the updated music audio data, the frequency components of the updated music audio data are identified through Fourier transform. The main frequency component of the updated music audio data is determined by counting the frequency times of each frequency component. At the same time, the frequency components are plotted in the frequency domain space to determine the phase information of each frequency component. Then, the single beat feature and the main beat feature of the updated music audio data are determined according to the frequency difference between the frequency component and the main frequency component, and thus the beat information of the updated music audio data is determined. Then, the initial period of the updated music audio data is identified, and the initial period is adjusted using the beat difference. Then, the core melody is identified by analyzing the melody of the updated music audio data. Further, the data text of the updated music audio data is identified, and the data emotion of the updated music audio data is trained by performing emotion annotation on the data text. Finally, the beat period feature and the melody emotion feature of the updated music audio data are generated based on the known information. In this way, in-depth analysis of the updated music audio data can be carried out, and all information of the updated music audio data is analyzed in detail, generating accurate beat period features and melody emotion features, ensuring the accuracy of subsequent clustering analysis.

[0111] Embodiment 4

[0112] Based on the Embodiment 3, for the intelligent clustering system of music data based on audio feature analysis, the melody determination unit includes:

[0113] A melody decomposition subunit, configured to decompose the updated music audio data to obtain the pitch feature, volume feature, speech rate feature, duration feature, and pitch weight feature of the updated music audio data;

[0114] A numerical analysis subunit, configured to analyze the numerical grade of the corresponding feature value according to the numerical grading standard corresponding to each melody dimension, and combine the numerical grades to generate the combined melody information of the updated music audio data;

[0115] A melody reconstruction subunit, configured to construct a pure audio according to the combined melody information, and generate the core melody of the updated music audio data according to the pure melody of the pure audio.

[0116] In this example, the pitch feature represents the frequency feature of the updated music audio data;

[0117] In this example, the volume feature represents the intensity of the updated music audio data;

[0118] In this example, the speech rate feature represents the language speed of the updated music audio data;

[0119] In this example, the duration feature represents the duration of each syllable, word, or sentence in the updated music audio data;

[0120] In this example, the stress feature represents the emphasized part of the updated music audio data;

[0121] In this example, the numerical grading standard represents the grading standard corresponding to a melody dimension.

[0122] The working principle and beneficial effects of the above technical solution: When decomposing the melody of the updated music audio data, numerical analysis is performed on each of its dimensions to determine the combined melody information of the updated music audio data, and then a pure audio is constructed to identify the core melody. In this way, the core melody of the updated music audio data is determined, and the simple characteristics of pure audio notes are utilized to achieve melody recognition, thereby determining the core melody of the updated music audio data.

[0123] Embodiment 5

[0124] Based on Embodiment 1, for the intelligent clustering system of music data based on audio feature analysis, the clustering processing module includes:

[0125] A preliminary classification unit, configured to match a corresponding first set of music categories for the updated music audio data according to the beat period feature, match a corresponding second set of music categories for the updated music audio data according to the melody emotion feature, and determine the initial music category corresponding to the updated music audio data according to the intersection between the first set of music categories and the second set of music categories;

[0126] A clustering analysis unit, configured to respectively obtain the existing music audio data corresponding to each of the initial music categories, construct a category attribute graph corresponding to each of the initial music categories based on the music labels corresponding to each of the existing music audio data, input the updated music audio data into the category attribute graph for classification verification, obtain the matching nodes of each of the updated music audio data in each of the category attribute graphs, and screen out the music categories whose node distances between the matching nodes and the corresponding existing nodes are less than the specified distance;

[0127] A parameter generation unit, configured to generate category features corresponding to the music categories according to the category attribute graph, perform fitting matching between the beat period feature and the category features to generate the beat period parameters of the updated music audio data in the corresponding music categories, perform fitting matching between the melody emotion feature and the category features to generate the melody emotion parameters of the updated music audio data in the corresponding music categories, and perform data fusion on the beat period parameters and the melody emotion parameters to generate the clustering adaptation parameters of the updated music audio data in each of the music categories.

[0128] In this example, the initial music category represents the music category that matches the updated music audio data;

[0129] In this example, the category attribute graph represents a graph of the attribute relationships of all the music in this music category constructed with the music labels of the existing music audio data as nodes;

[0130] In this example, the specified distance is 1.

[0131] The working principle and beneficial effects of the above technical solution: First, separately match the first music category set for the updated music audio data according to the beat period feature, and at the same time separately use the melody emotion feature to match the second music category set for the updated music audio data. Determine the initial music category of the updated music audio data according to the intersection of the two music category sets, and then generate a category attribute graph for the music labels corresponding to the existing music audio data of each initial music category, and then classify and verify the updated music audio data, and screen out the music categories that meet the node distance requirements. Then further match the beat period feature of the updated music audio data with the category feature of the music category and match the melody emotion feature with the category feature to determine the beat period parameter and melody emotion parameter of the updated music audio data. Generate the clustering adaptation parameter of the updated music audio data in this way. During the process of generating the parameter, both the beat period feature and the melody emotion feature are matched with the category feature of the music category, improving the matching accuracy, avoiding contingency, and ensuring the clustering effect.

[0132] Embodiment 6

[0133] Based on Embodiment 5, for the music data intelligent clustering system based on audio feature analysis, the preliminary classification unit is further used for:

[0134] When there is no intersection between the first music category set and the second music category set, determine that the updated music audio data is abnormal and generate a re-acquisition instruction;

[0135] Use the re-acquisition instruction to control the data acquisition module to re-acquire the updated music audio data.

[0136] The working principle and beneficial effects of the above technical solution: When the initial music category cannot be generated, control the data acquisition module to re-acquire the updated music audio data, ensuring the integrity and accuracy of the updated music audio data.

[0137] Embodiment 7

[0138] Based on Embodiment 1, for the music data intelligent clustering system based on audio feature analysis, the clustering optimization module includes:

[0139] A clustering verification unit, which is used to iteratively cluster the updated music audio data with the corresponding music categories based on the clustering adaptation parameters, so as to obtain the data similarity features between the updated music audio data and the existing music audio data in each music category;

[0140] A clustering optimization unit, which is used to count several data similarity features corresponding to each music category to determine the clustering similarity between the updated music audio data and each music category, and determine several target music categories corresponding to each updated music audio data based on the clustering similarity;

[0141] A division execution unit, which is used to respectively screen out the relevant music audio data with a clustering similarity higher than the standard similarity with the updated music audio data in each target music category, set a recommendation guide for the updated music audio data, and divide the updated music audio data into the corresponding target music categories respectively.

[0142] In this example, iterative clustering refers to the process of adjusting the characteristics of the updated music audio data using the clustering adaptation parameters and then performing multiple matches with the music categories;

[0143] In this example, the data similarity feature refers to the data similarity feature between the existing music audio data and the updated music audio data.

[0144] The working principle and beneficial effects of the above technical solution: By iteratively clustering the updated music audio data and the music categories to determine the data similarity between the updated music audio data and each existing music in the music category, thereby determining the clustering similarity between the updated music audio data and each music category, and thus determining the target music category, and providing recommendation guidance for the updated music audio data in the target music category, improving the accurate exposure of the updated music audio data and the effect of keeping the music software up with the times.

[0145] Embodiment 8

[0146] Based on Embodiment 7, the intelligent clustering system for music data based on audio feature analysis further includes:

[0147] After updating the updated music audio data into the corresponding target music category, regard the updated music audio data as existing music audio data.

[0148] The working principle and beneficial effects of the above technical solution: Regarding the updated music audio data as existing music audio data after completing the update of the target music category is convenient for the next clustering.

[0149] Embodiment 9

[0150] This embodiment provides an intelligent clustering method for music data based on audio feature analysis, as follows Figure 2 shown, including:

[0151] Step 1: Collect updated music audio data of the music software;

[0152] Step 2: Identify the beat period feature and the melody emotion feature of the updated music audio data;

[0153] Step 3: Identify the music category corresponding to the updated music audio data according to the beat period feature and the melody emotion feature, and construct a clustering adaptation parameter according to the category feature corresponding to each music category;

[0154] Step 4: Determine the clustering similarity between the updated music audio data and each music category according to the clustering adaptation parameter, and divide the updated music audio data into the corresponding target music category.

[0155] In this example, the beat period feature represents the rhythm basis and repetition period of the updated music audio data;

[0156] In this example, the melody emotion feature represents the melody composed of the notes of the updated music audio data and the emotion it expresses;

[0157] In this example, the clustering adaptation parameter represents the parameter generated when the updated music audio data is clustered into a music category and adapts to this music category;

[0158] In this example, the clustering similarity represents the similarity between the updated music audio data and the existing music audio data in the music category;

[0159] In this example, the number of target music categories can be one or more.

[0160] Working principle and beneficial effects of the above technical solution: When the music audio data is updated in the music software, the data is collected in a timely manner and the beat period characteristics and melody emotion characteristics of the data are identified. Then, according to the beat period characteristics and melody emotion characteristics, the corresponding classification is matched for the updated music audio data, and the corresponding clustering adaptation parameters are constructed. Further, the parameters are used to optimize the clustering of the updated music audio data, and the target music category that the updated music audio data conforms to is determined. In this way, the updated music audio data in the music software can be captured in a timely manner, and then its necessary characteristics are identified, and then clustering is performed. In order to improve the accuracy of clustering, the clustering adaptation parameters between the updated music audio data and different music categories are determined, and then the parameters are used to deeply analyze the clustering similarity between the two, so as to determine the target music category of the updated music audio data, improve the accuracy of the traditional clustering technology, and can assign a suitable category to the updated music audio data, which is convenient for listeners to select their favorite music category and improves the listeners' love for the music software.

[0161] Example 10

[0162] Based on the music data intelligent clustering method based on audio feature analysis in Example 9, step 2 includes:

[0163] Step 21: Perform Fourier transform on the updated music audio data to obtain a plurality of frequency components, count the frequency times corresponding to each frequency component, determine the main frequency component of the updated music audio data, and respectively plot each frequency component in the frequency domain space to obtain the phase information corresponding to each frequency component;

[0164] Step 22: Determine a plurality of single beat characteristics of the updated music audio data based on the frequency difference between each frequency component and the main frequency component, determine the main beat characteristic of the updated music audio data based on the main frequency component, and use each single beat characteristic to correct the main beat characteristic to obtain the beat information of the updated music audio data;

[0165] Step 23: Generate an initial period according to the main phase information corresponding to the main frequency component, decompose the beat information by using the initial period to obtain a plurality of decomposed beats, and adjust the period range of the initial period according to the beat difference between different decomposed beats;

[0166] Step 24: Decompose the updated music audio data into melodies to obtain the melody characteristics corresponding to the updated music audio data in different melody dimensions, and determine the core melody of the updated music audio data according to the characteristic values corresponding to each melody characteristic;

[0167] Step 25: Use natural language processing technology to perform text recognition on the updated music audio data to obtain corresponding data text, use a preset sentiment dictionary to perform sentiment annotation on the data text, determine the sentiment tendency of the updated music audio data, and use the random forest method to train the sentiment tendency to obtain the data sentiment of the updated music audio data;

[0168] Step 26: Generate the beat period feature of the updated music audio data according to the beat information and the period range, and generate the melody sentiment feature of the updated music audio data according to the core melody and the data sentiment.

[0169] In this example, the frequency component represents the component regarding different frequencies included in the updated music audio data;

[0170] In this example, the frequency count represents the number of times a frequency component appears;

[0171] In this example, the main frequency component represents the frequency component with the most occurrences in the updated music audio data;

[0172] In this example, the phase information represents the waveform presented by the frequency component;

[0173] In this example, the frequency difference represents the phase shift between the frequency component and the main frequency component;

[0174] In this example, the single beat feature represents the beat presented by a frequency component;

[0175] In this example, the main beat feature represents the beat presented by the main frequency component;

[0176] In this example, the beat information represents the continuous beats presented by the updated music audio data;

[0177] In this example, the process of performing beat correction on the main beat feature is: the process of using the single beat feature to correct the notes of the main beat feature;

[0178] In this example, the initial period represents the period deduced according to the main phase information;

[0179] In this example, the period range represents the period length range of the initial period;

[0180] In this example, the melody dimension includes the pitch dimension, volume dimension, speech rate dimension, duration dimension, and stress dimension;

[0181] In this example, the core melody represents the melody presented multiple times by the updated music audio data;

[0182] In this example, the sentiment annotation represents the process of matching corresponding sentiment to the data text;

[0183] In this example, a preset sentiment dictionary contains a number of words, and each word is set with a sentiment value. The sentiment tendency of a text is identified by counting the sentiment values contained in the text.

[0184] In this example, the data sentiment represents the sentiment presented by the updated music audio data.

[0185] The working principle and beneficial effects of the above technical solution: In order to be able to identify the various characteristics of the updated music audio data, the frequency components of the updated music audio data are identified through Fourier transform, the main frequency component of the updated music audio data is determined by counting the frequency times of each frequency component, and at the same time, the frequency components are plotted in the frequency domain space to determine the phase information of each frequency component. Then, the single beat feature and the main beat feature of the updated music audio data are determined according to the frequency difference between the frequency component and the main frequency component, and then the beat information of the updated music audio data is determined. Then, the initial period of the updated music audio data is identified, and the initial period is adjusted by using the beat difference. Then, the core melody is identified by analyzing the melody of the updated music audio data, and the data text of the updated music audio data is further identified. The data text is sentimentally annotated to train the data sentiment of the updated music audio data. Finally, the beat period feature and the melody sentiment feature of the updated music audio data are generated according to the known information. In this way, the updated music audio data can be deeply analyzed, and all the information of the updated music audio data is analyzed in detail, generating accurate beat period features and melody sentiment features, ensuring the accuracy of subsequent clustering analysis.

[0186] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these changes and modifications.

Claims

1. An intelligent music data clustering system based on audio feature analysis, characterized in that: include: A data acquisition module, used for acquiring updated music audio data of the music software; A feature extraction module, used to identify the beat cycle characteristics and melody emotion characteristics of the updated music audio data; A clustering processing module, used for identifying the music category corresponding to the updated music audio data according to the beat cycle characteristics and the melody emotion characteristics, and constructing a clustering adaptation parameter according to the category characteristics corresponding to each of the music categories; A clustering optimization module is used to determine the clustering similarity between the updated music audio data and each of the music categories according to the clustering adaptation parameters, and to divide the updated music audio data into corresponding target music categories.

2. The music data intelligent clustering system based on audio feature analysis according to claim 1, characterized in that: The data acquisition module comprises: A data capture unit, used to capture real-time complete update data of the music software; The data screening unit is used to perform duration analysis and type analysis on each of the real-time complete update data, and screen the updated music audio data that meets the specified duration and specified type.

3. The music data intelligent clustering system based on audio feature analysis as claimed in claim 1, characterized in that: The feature extraction module comprises: A data processing unit, configured to perform Fourier transform on the updated music audio data to obtain a plurality of frequency components, count the frequency times corresponding to each of the frequency components, determine the main frequency components of the updated music audio data, and plot each of the frequency components in the frequency domain space to obtain phase information corresponding to each of the frequency components; A beat determination unit, configured to determine a plurality of single beat features of the updated music audio data based on a frequency difference between each of the frequency components and the main frequency component, determine a main beat feature of the updated music audio data based on the main frequency component, perform beat correction on the main beat feature using each of the single beat features, and obtain beat information of the updated music audio data; A period determination unit, configured to generate an initial period according to the main phase information corresponding to the main frequency component, decompose the beat information using the initial period to obtain a plurality of decomposed beats, and adjust the period range of the initial period according to the beat difference between different decomposed beats; a melody determination unit, configured to perform melody decomposition on the updated music audio data, obtain melody features corresponding to the updated music audio data under different melody dimensions, and determine a core melody of the updated music audio data according to a feature value corresponding to each melody feature; An emotion determination unit is used to perform text recognition on the updated music audio data using natural language processing technology to obtain corresponding data text, perform emotion annotation on the data text using a preset emotion dictionary to determine the emotion tendency of the updated music audio data, and train the emotion tendency using a random forest method to obtain data emotion of the updated music audio data; A feature determination unit is used to generate a beat period feature of the updated music audio data according to the beat information and the period range, and to generate a melody emotion feature of the updated music audio data according to the core melody and the data emotion.

4. The music data intelligent clustering system based on audio feature analysis as claimed in claim 3, characterized in that: The melody determination unit comprises: A melody decomposition subunit, used for performing melody decomposition on the updated music audio data to obtain pitch features, volume features, speech speed features, duration features and tone weight features of the updated music audio data; a numerical analysis subunit, for analyzing the numerical levels of the corresponding characteristic numerical values ​​according to the numerical grading standard corresponding to each melody dimension, combining the numerical levels, and generating combined melody information of the updated music audio data; The melody reconstruction subunit is used to construct pure audio according to the combined melody information, and generate the core melody of the updated music audio data according to the pure melody of the pure audio.

5. The music data intelligent clustering system based on audio feature analysis as claimed in claim 1, characterized in that: The cluster processing module comprises: A preliminary classification unit, configured to match a corresponding first music category set for the updated music audio data according to the beat cycle feature, match a corresponding second music category set for the updated music audio data according to the melody emotion feature, and determine a preliminary music category corresponding to the updated music audio data according to an intersection between the first music category set and the second music category set; A cluster analysis unit, for respectively obtaining existing music audio data corresponding to each of the initial music categories, constructing a category attribute graph corresponding to each of the initial music categories based on the music tags corresponding to each of the existing music audio data, inputting the updated music audio data into the category attribute graph for classification verification, obtaining matching nodes of each of the updated music audio data in each of the category attribute graphs, and screening music categories whose node distance between the matching node and the corresponding existing node is less than a specified distance; A parameter generating unit is used to generate category features corresponding to the music category according to the category attribute graph, fit and match the beat period features with the category features to generate beat period parameters of the updated music audio data in the corresponding music category, fit and match the melody emotion features with the category features to generate melody emotion parameters of the updated music audio data in the corresponding music category, perform data fusion of the beat period parameters and the melody emotion parameters to generate clustering adaptation parameters of the updated music audio data in each of the music categories.

6. The music data intelligent clustering system based on audio feature analysis as claimed in claim 5, characterized in that: The preliminary taxonomic units are also used for: When the first music category set and the second music category set do not contain an intersection, determining that the updated music audio data is abnormal, and generating a re-acquisition instruction; The re-acquisition instruction is used to control the data acquisition module to re-acquire the updated music audio data.

7. The music data intelligent clustering system based on audio feature analysis according to claim 1, characterized in that: The clustering optimization module comprises: A clustering verification unit, configured to iteratively cluster the updated music audio data with the corresponding music categories based on the clustering adaptation parameter, and obtain data similarity features between the updated music audio data and the existing music audio data in each of the music categories; A clustering optimization unit, used for determining a clustering similarity between the updated music audio data and each of the music categories by counting a number of data similarity features corresponding to each of the music categories, and determining a number of target music categories corresponding to each of the updated music audio data based on the clustering similarity; A division execution unit is used to screen relevant music audio data whose clustering similarity with the updated music audio data is higher than the standard similarity in each of the target music categories, set a recommendation guide for the updated music audio data, and divide the updated music audio data into corresponding target music categories.

8. The music data intelligent clustering system based on audio feature analysis as claimed in claim 7, characterized in that: Also includes: After the updated music audio data is updated to the corresponding target music category, the updated music audio data is regarded as existing music audio data.

9. An intelligent clustering method for music data based on audio feature analysis, characterized in that: include: Step 1: Collect updated music audio data of music software; Step 2: Identify the beat cycle characteristics and melody emotion characteristics of the updated music audio data; Step 3: Identify the music category corresponding to the updated music audio data according to the beat cycle feature and the melody emotion feature, and construct a clustering adaptation parameter according to the category feature corresponding to each of the music categories; Step 4: Determine the clustering similarity between the updated music audio data and each of the music categories based on the clustering adaptation parameters, and classify the updated music audio data into the corresponding target music category.

10. The music data intelligent clustering method based on audio feature analysis according to claim 9, characterized in that: The step 2 comprises: Step 21: Perform Fourier transform on the updated music audio data to obtain a plurality of frequency components, count the frequency times corresponding to each of the frequency components, determine the main frequency components of the updated music audio data, and plot each of the frequency components in the frequency domain space to obtain the phase information corresponding to each of the frequency components; Step 22: determining a plurality of single beat features of the updated music audio data based on the frequency difference between each of the frequency components and the main frequency component, determining the main beat feature of the updated music audio data based on the main frequency component, performing beat correction on the main beat feature using each of the single beat features, and obtaining beat information of the updated music audio data; Step 23: Generate an initial cycle according to the main phase information corresponding to the main frequency component, use the initial cycle to decompose the beat information to obtain a plurality of decomposed beats, and adjust the period range of the initial cycle according to the beat difference between different decomposed beats; Step 24: performing melody decomposition on the updated music audio data to obtain melody features corresponding to the updated music audio data under different melody dimensions, and determining the core melody of the updated music audio data according to the feature value corresponding to each melody feature; Step 25: using natural language processing technology to perform text recognition on the updated music audio data to obtain corresponding data text, using a preset emotional dictionary to perform emotional annotation on the data text to determine the emotional tendency of the updated music audio data, and using a random forest method to train the emotional tendency to obtain the data emotion of the updated music audio data; Step 26: Generate the beat period characteristics of the updated music audio data according to the beat information and the period range, and generate the melody emotion characteristics of the updated music audio data according to the core melody and the data emotion.

Images