Music intelligent recommendation and music data sharing method and system

Abstract

The invention discloses a music intelligent recommendation and music data sharing method and system, and particularly relates to the technical field of music recommendation. Music playing data, search data and user interaction data of a user on a plurality of music platforms are integrated, a global music preference model is constructed, a personalized recommendation list is generated in real time, the recommendation accuracy is dynamically evaluated and optimized, and through a safe and efficient data sharing and migration mechanism, the user experience is improved. Sharing the music preference data of the user among different platforms; meanwhile, the synchronization efficiency and the network stability in the data transmission process are analyzed through fuzzy logic, the safety and the stability of data sharing are ensured, the problems that scattered data are difficult to integrate, recommendation is not accurate and data sharing is not safe are effectively solved, and the accuracy and the user experience of the music recommendation system are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of music recommendation, and particularly relates to a method and system for intelligent music recommendation and music data sharing. Background Art

[0002] Intelligent music recommendation refers to using artificial intelligence and big data technologies to recommend music content that suits the user's taste based on the user's personalized data such as music preferences, listening history, emotional state, etc. This recommendation system analyzes a large amount of music data, identifies user preferences, and automatically generates playlists or recommends new songs, albums, and artists, thereby enhancing the user's music experience.

[0003] Music data sharing refers to the sharing of music-related data between users or platforms, such as the user's play history, favorite tracks, evaluations, and comments. The sharing of these data can be used for social interaction, helping users discover new music, and can also be used to improve music recommendation algorithms to make recommendations more accurate and personalized. In addition, music data sharing can also provide valuable market insights and feedback for artists, record companies, and developers.

[0004] The existing technologies have the following deficiencies:

[0005] In the current music recommendation system, the user's music preferences and behavior data are often scattered in different music platforms, making it difficult to effectively integrate and analyze. This results in the recommendation system being unable to accurately capture the user's overall music preferences, thereby affecting the accuracy of the recommendation results. In addition, the existing music data sharing methods lack security, and users cannot efficiently and securely share and migrate music data between different platforms. Summary of the Invention

[0006] The purpose of the present invention is to provide a method and system for intelligent music recommendation and music data sharing to solve the deficiencies in the background art.

[0007] To achieve the above purpose, the present invention provides the following technical solution: A method for intelligent music recommendation and music data sharing, comprising the following steps:

[0008] S1: Obtain the music preference data of the user on several music platforms, where the music preference data includes music play data, search data, and user interaction data;

[0009] S2: Construct a global music preference model of the user according to the music preference data of the user on each platform, and generate a real-time music play recommendation list through the global music preference model;

[0010] S3: Evaluate the accuracy of the global music preference model based on the change in the playback rate of the music playback recommendation list by the user and the similarity degree between the music searched by the user and the music recommended in the music playback recommendation list;

[0011] S4: According to the evaluation results, divide the accuracy of the global music preference model into different levels, divide it into high accuracy level, general accuracy level and low accuracy level, and make corresponding adjustments to the music playback recommendation list;

[0012] S5: According to the adjusted music playback recommendation list, share and migrate the music preference data of the user between different music platforms, and monitor it in real time to obtain the data synchronization efficiency fluctuation state during the data transmission process and the stability of the communication network;

[0013] S6: Analyze the data synchronization efficiency fluctuation state during the data transmission process and the stability of the communication network through fuzzy logic, evaluate the security of the music preference data sharing process, and optimize the music preference data sharing process according to the evaluation results.

[0014] Preferably, in S3, based on the change in the playback rate of the music playback recommendation list by the user and the similarity degree between the music searched by the user and the music recommended in the music playback recommendation list, evaluate the accuracy of the global music preference model, specifically:

[0015] Generate a track jump density fluctuation index according to the change in the playback rate of the music playback recommendation list by the user. The method for obtaining the track jump density fluctuation index is as follows:

[0016] Calculate the track jump density of the user within a time window and record it as a time series y1, y2, …, yt, …, yT, where yt represents the track jump density at time t. Divide the jump behavior into three states: low jump state (L), medium jump state (M), high jump state (H); Discretize the time series yt into an observation sequence O = {o1, o2, …, oT}, where ot represents the discrete observation value at time t; The initial state probability vector π = {π1, π2, …, πN}: represents the probability that the system belongs to state i at the initial moment, where N is the number of states, and the state transition matrix A = {aij}: represents the probability of transitioning from state i to state j, where, ; The observation probability matrix B = {bj(o)}: represents the probability of observing ot = o in state j;

[0017] Train it using the Baum-Welch algorithm, update the model parameters λ=(π, A, B), adjust the parameters by maximizing the likelihood function of the observation sequence, and iterate repeatedly on the given training data until the parameters converge; use the Viterbi algorithm to find the target state sequence S={s1, s2, …, sT}, that is, the user's jump behavior pattern in different time periods, and the expression is: ; In the formula, is the optimal path probability in the Viterbi algorithm, represents the optimal path probability that the system is in state i at time . Trace back the optimal state sequence from the final time T, and calculate the transition frequency between the high jump density state (H) and the low jump density state (L), that is, calculate the track jump density fluctuation index, and the expression is: ; In the formula, TSD is the track jump density fluctuation index.

[0018] Preferably, in S3, generate a rhythm matching rate deviation index according to the similarity degree between the user's searched music and the recommended music in the music play recommendation list. The method for obtaining the rhythm matching rate deviation index is:

[0019] Collect the rhythm data of the user's searched music, including the beat rate, and at the same time collect the BPM value of each music in the recommended music list; perform standardization processing on the BPM values to generate two standardized BPM sequences, one corresponding to the user's searched music and the other corresponding to the recommended music; for each music searched by the user and each music in the recommended list, calculate their rhythm matching rate , and the expression is: ; In the formula, is the BPM value of the music searched by the user, is the BPM value of the recommended music. Average the RMR values of all music pairs to obtain the overall rhythm matching rate , and the expression is: ; In the formula, n is the number of pairs of the user's searched music and the recommended music. Calculate the rhythm matching deviation of each pair of music, and the expression is: ; Average the matching deviations of all music pairs to obtain the rhythm matching rate deviation index RMRD, and the expression is: .

[0020] Preferably, in S3, evaluate the accuracy of the global music preference model, specifically:

[0021] Convert the track jump density fluctuation index and the rhythm matching rate deviation index into a first feature vector, and use the first feature vector as the input of a machine learning model. The machine learning model takes predicting the accuracy value label of the global music preference model for each group of first feature vectors as the prediction target, and takes minimizing the sum of the prediction errors for the accuracy value labels of all global music preference models as the training target. Train the machine learning model until the sum of the prediction errors reaches convergence and then stop the model training. Determine the accuracy value of the global music preference model according to the model output result, where the machine learning model is a polynomial regression model.

[0022] Preferably, in S4, divide the accuracy of the global music preference model into different levels, specifically:

[0023] Compare the obtained accuracy value of the global music preference model with the gradient standard thresholds. The gradient standard thresholds include a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold. Compare the accuracy value of the global music preference model with the first standard threshold and the second standard threshold respectively;

[0024] If the accuracy value of the global music preference model is greater than the second standard threshold, it indicates that the accuracy of the global music preference model is high. Divide it into the high-accuracy level and generate a model high-accuracy signal; if the accuracy value of the global music preference model is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it indicates that the accuracy of the global music preference model is average. Divide it into the medium-accuracy level and generate a model medium-accuracy signal; if the accuracy value of the global music preference model is less than the first standard threshold, it indicates that the accuracy of the global music preference model is low. Divide it into the low-accuracy level and generate a model low-accuracy signal.

[0025] Preferably, in S5, perform real-time monitoring to obtain the data synchronization efficiency fluctuation state during the data transmission process and the stability of the communication network, specifically:

[0026] Generate a transmission delay jitter index according to the data synchronization efficiency fluctuation state during the data transmission process. The method for obtaining the transmission delay jitter index is as follows:

[0027] Collect the delay time series of each data packet during the data transmission process, perform normalization processing on the delay data, and the normalized delay time series , where n is the number of the data packet; select the wavelet function ψ(s) to decompose the delay data, s is the time, determine the number of layers J of the wavelet decomposition, and use the discrete wavelet transform to decompose the delay signal into the approximation coefficients and detail coefficients at different scales. The expression is: ; is the coefficient at the j-th layer scale, is the corresponding wavelet basis function, and the detail coefficients of each layer are extracted , and the expression is: ; The detail coefficients of each layer , , ... , where represents the detail coefficient of the j-th layer, and the energy of the detail coefficient of each layer is calculated , and the expression is: ; where is the k-th detail coefficient of the j-th layer, is the corresponding energy, and the energy of the detail coefficients of each layer is , , ... ; By synthesizing the energies of the detail coefficients of all layers, the total jitter energy is obtained , and the expression is: ; By normalizing the total energy, the transmission delay jitter index is obtained , and the expression is: ; where N is the total length of the delay signal.

[0028] Preferably, according to the stability of the communication network during the data transmission process, the retransmission aggregation index is generated, and the method for obtaining the retransmission aggregation index is as follows:

[0029] Cluster the retransmission events according to the time ti and the network location pi, and divide the retransmission events into several clustering clusters C1, C2,..., Cm, where m is the number of clustering clusters, and count the number of retransmission events in each clustering cluster , where j = 1, 2,..., m; the retransmission event clusters after clustering and the number of retransmissions in each cluster; calculate the weight of each clustering cluster , that is, the ratio of the number of retransmissions in each cluster to the total number of retransmissions, and the expression is: ; In the formula, n is the total number of retransmission events, and calculate the sum of the squares of the weights of all clusters to represent the aggregation degree, that is, calculate the retransmission aggregation index, and the expression is: ; In the formula, is the retransmission aggregation index.

[0030] Preferably, in S6, the fuzzy logic is used to analyze the data synchronization efficiency fluctuation state and the stability of the communication network during the data transmission process, and evaluate the security of the music preference data sharing process, specifically:

[0031] Take the transmission delay jitter index TD and the retransmission aggregation index MR as the input items of the fuzzy logic, and take the security value WD of the music preference data sharing process as the output item of the fuzzy logic;

[0032] Define fuzzy sets for each input variable TD and MR, and also define a fuzzy set for the output variable WD;

[0033] Define membership degrees for each fuzzy set, that is, define membership functions for the retransmission aggregation degree index MR and the output security value WD;

[0034] For specific TD and MR values, calculate their membership degrees in each fuzzy set;

[0035] The fuzzy membership degrees of the input variables TD and MR serve as the basis for fuzzy logic rules;

[0036] Design fuzzy rules to map the input variables to the output variable;

[0037] Construct a fuzzy rule table to map all TD and MR combinations to the corresponding WD output;

[0038] According to the current TD and MR values, activate the corresponding fuzzy rules;

[0039] Perform inference on all activated rules, and calculate the membership degrees of the output variable WD in each fuzzy set; use the min-max inference method to synthesize the outputs of the activated rules, and the expression is: ; In the formula, and are input membership functions, is the output membership function;

[0040] Use the Centroid Method to defuzzify the output membership function to obtain the exact security value WD, and the expression is: ; Among them, is the corresponding membership degree.

[0041] Preferably, according to the calculated security value, judge the security of the current data sharing process, compare the obtained security value with the security value reference threshold. If the security value is greater than or equal to the security value reference threshold, no warning signal is generated at this time and no processing is required; if the security value is less than the security value reference threshold, a warning signal is generated at this time and optimization measures are taken.

[0042] The present invention also provides a music intelligent recommendation and music data sharing system, including a data acquisition module, a global music preference model construction module, an accuracy evaluation module, an adjustment module, a data sharing module, and a security analysis module;

[0043] Data acquisition module: Acquire music preference data of users on several music platforms, and the music preference data includes music play data, search data, and user interaction data;

[0044] Global Music Preference Model Construction Module: Construct a global music preference model for the user based on the music preference data of the user on various platforms, and generate a real-time music play recommendation list through the global music preference model;

[0045] Accuracy Evaluation Module: Evaluate the accuracy of the global music preference model according to the change in the play rate of the music play recommendation list by the user and the similarity degree between the music searched by the user and the music recommended in the music play recommendation list;

[0046] Adjustment Module: According to the evaluation results, divide the accuracy of the global music preference model into different levels, divide it into high accuracy level, general accuracy level and low accuracy level, and make corresponding adjustments to the music play recommendation list;

[0047] Data Sharing Module: According to the adjusted music play recommendation list, share and migrate the music preference data of the user between different music platforms, and monitor it in real time to obtain the data synchronization efficiency fluctuation state during the data transmission process and the stability of the communication network;

[0048] Security Analysis Module: Analyze the data synchronization efficiency fluctuation state and the stability of the communication network during the data transmission process through fuzzy logic, evaluate the security of the music preference data sharing process, and optimize the music preference data sharing process according to the evaluation results.

[0049] In the above technical solution, the technical effects and advantages provided by the present invention:

[0050] 1. By integrating the music play data, search data and user interaction data of the user on multiple music platforms, the present invention constructs a global music preference model and generates a personalized recommendation list in real time, dynamically evaluates and optimizes the accuracy of the recommendation, and realizes more accurate music recommendation. At the same time, by analyzing the data synchronization efficiency fluctuation state and the stability of the communication network during the data transmission process through fuzzy logic, the security of the music data sharing process is effectively evaluated and improved, ensuring the efficient and secure migration of user data between different platforms.

[0051] 2. The present invention solves the security challenges in the music data sharing process through an innovative security evaluation and optimization mechanism. Overall, the present invention improves the accuracy of the recommendation system and the user experience, enhances the stability and reliability of cross-platform data sharing, and enables users to enjoy a consistent music experience between multiple platforms. Description of the Drawings

[0052] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for use in the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other accompanying drawings can also be obtained based on these drawings.

[0053] Figure 1 It is a flowchart of the method of the present invention.

[0054] Figure 2 It is a system module diagram of the present invention. Specific embodiments

[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0056] Embodiment 1. Please refer to Figure 1 As shown, a music intelligent recommendation and music data sharing method in this embodiment includes the following steps:

[0057] S1: Obtain the music preference data of the user on several music platforms. The music preference data includes music play data, search data, and user interaction data;

[0058] S2: Construct a global music preference model of the user according to the music preference data of the user on each platform, and generate a real-time music play recommendation list through the global music preference model;

[0059] S3: Evaluate the accuracy of the global music preference model according to the change in the play rate of the user for the music play recommendation list and the similarity degree between the music searched by the user and the music recommended in the music play recommendation list;

[0060] S4: According to the evaluation results, divide the accuracy of the global music preference model into different levels, divide it into high accuracy level, general accuracy level, and low accuracy level, and make corresponding adjustments to the music play recommendation list;

[0061] S5: According to the adjusted music play recommendation list, share and migrate the music preference data of the user between different music platforms, and monitor it in real time to obtain the data synchronization efficiency fluctuation state during the data transmission process and the stability of the communication network;

[0062] S6: Analyze the data synchronization efficiency fluctuation state during data transmission and the stability of the communication network through fuzzy logic, evaluate the security of the music preference data sharing process, and optimize the music preference data sharing process according to the evaluation results.

[0063] Among them, in S1, obtain the music preference data of the user on several music platforms, and the music preference data includes music play data, search data and user interaction data. Specifically:

[0064] Obtain user authorization: Through OAuth 2.0 or other authentication protocols, obtain the user's data access authorization for each music platform. The user needs to log in on each platform and grant the application the permission to access their music data. Ensure the legitimacy of the user's identity and prevent unauthorized access. Identity confirmation can be carried out through methods such as SMS verification codes and email verification.

[0065] Integrate the APIs of each platform: Each music platform usually provides an API to access user data. By integrating these APIs, data requests can be sent to each platform.

[0066] Send data requests: Through API calls, request the user data on each platform respectively, including music play data, search data and interaction data. These requests usually include parameters such as user ID and time range to ensure that accurate and relevant data is obtained.

[0067] Data reception: Receive the response data from each platform. This data is usually returned in JSON or XML format.

[0068] Data format standardization: Since the data formats returned by different platforms may be different, first standardize this data, unify the data format and field names to facilitate subsequent data processing and analysis.

[0069] Initial data cleaning: Remove invalid data (such as duplicates, null values), correct format errors, and ensure the quality of the data.

[0070] Data storage: Store the standardized and cleaned data in a centralized database (such as an SQL database or a NoSQL database), and establish corresponding data tables or collections according to different data types (play data, search data, interaction data).

[0071] Data indexing and structuring: Index the stored data and structure it according to dimensions such as time, user ID, and platform to improve the efficiency of data query.

[0072] S2: Construct a global music preference model for the user based on the music preference data of the user on each platform, and generate a real-time music play recommendation list through the global music preference model.

[0073] Integrate the music preference data of users on various platforms (including music play data, search data, user interaction data, etc.) to form a unified data set.

[0074] Merge the data of the same user on different platforms, remove duplicates, and generate a user data set containing all platform behaviors.

[0075] Feature extraction: Extract key features from the integrated data to provide a basis for building a global music preference model. The key features include play features, search features, and interaction features, specifically:

[0076] Play features: Extract the music genres, artists, albums, tracks, etc. that the user plays most frequently.

[0077] Search features: Analyze the keywords, music genres, etc. that appear frequently in the user's search history.

[0078] Interaction features: Extract the behavior features of the user's likes, comments, collections, shares, etc. on the music platform.

[0079] A structured data set containing various music preference features of users provides input for building a global music preference model.

[0080] Combine the data features of different platforms to build a global music preference model for users. Assign weights to the data features from different platforms, and the weights can be determined based on factors such as the usage frequency of the platform, the user's activity, and the integrity of the data.

[0081] Use feature fusion technology to integrate the features of each platform. For example, perform weighted averaging on similar features (such as the same type of music preferences on different platforms) to generate a unified preference feature vector.

[0082] Model selection and training: According to the fused feature data, select a suitable model to describe the user's global music preference. Make recommendations based on the user's music content preferences (such as music genres, artist styles, etc.). Combine collaborative filtering and content recommendation models to generate a more accurate global preference model. Use historical data to train the model and evaluate the accuracy of the model through methods such as cross-validation.

[0083] Maintain the real-time and dynamic nature of the recommendation list to ensure that the recommendation results can timely reflect the user's latest preferences. Real-time collect the user's latest behavior data (such as current play, search, interaction, etc.) and update the global music preference model. Use online learning algorithms to enable the model to gradually adapt to changes in user preferences.

[0084] Generate a personalized music recommendation list for the user at the current moment. Based on the global music preference model, filter out a candidate music set from the music library that highly matches the user's preferences. Score the candidate music, and the scoring criteria can include the degree of match with the user's preferences, the popularity of the music, the user's current situation, etc.

[0085] Apply business rules to filter out inappropriate recommendations (such as music the user has already listened to, currently unavailable content, etc.), and ensure the diversity of the recommendation list (avoid recommending too many music of the same type).

[0086] Push the generated recommendation list to the user for selection and playback. Convert the scored and sorted music recommendation list into a user interface-friendly format and update it in real-time and display it on the user's recommendation page. Output a real-time, personalized music playback recommendation list that conforms to the user's current music preferences and behavioral characteristics.

[0087] S3: Evaluate the accuracy of the global music preference model according to the change in the playback rate of the music playback recommendation list by the user and the similarity degree between the music searched by the user and the music recommended in the music playback recommendation list.

[0088] Generate a track jump density fluctuation index according to the change in the playback rate of the music playback recommendation list by the user. The method for obtaining the track jump density fluctuation index is as follows:

[0089] Calculate the track jump density of the user within a time window, for example, the number of jumps per minute or per hour divided by the total number of plays. Denote it as the time series y1, y2, …, yt, …, yT, where yt represents the track jump density at time t. Classify the jump behavior into three states: low jump state (L), medium jump state (M), and high jump state (H). These states reflect the user's behavioral patterns at different time periods.

[0090] Discretize the time series yt into an observation sequence O = {o1, o2, …, oT}, where ot represents the discrete observation value at time t (such as the interval of jump density).

[0091] Initial state probability vector π = {π1, π2, …, πN}: Represents the probability that the system belongs to state i at the initial moment, where N is the number of states.

[0092] State transition matrix A = {aij}: Represents the probability of transitioning from state i to state j, where .

[0093] Observation probability matrix B = {bj(o)}: Represents the probability of observing ot = o in state j.

[0094] Train it using the Baum-Welch algorithm, update the model parameters λ = (π, A, B), adjust the parameters by maximizing the likelihood function of the observation sequence, and iterate repeatedly on the given training data until the parameters converge.

[0095] Use the Viterbi algorithm to find the target state sequence S = {s1, s2, …, sT}, that is, the user's jump behavior pattern in different time periods. The expression is: ; In the formula, is the optimal path probability in the Viterbi algorithm, represents the optimal path probability that the system is in state i at time t. Trace back the optimal state sequence from the final time T, and calculate the transition frequency between the high jump density state (H) and the low jump density state (L), that is, calculate the track jump density fluctuation index. The expression is: ; In the formula, TSD is the track jump density fluctuation index.

[0096] The larger the track jump density fluctuation index, the more frequently the user switches between different states, indicating high instability or dissatisfaction with the recommended tracks. This shows that the global music preference model fails to accurately capture the user's true preferences, resulting in a mismatch between the recommended content and the user's needs. Therefore, the larger the track jump density fluctuation index, usually the lower the accuracy of the global music preference model, and the model needs to be optimized or adjusted to improve the relevance of the recommendation and user satisfaction.

[0097] Generate a rhythm matching rate deviation index according to the similarity degree between the music searched by the user and the music in the music play recommendation list. The method for obtaining the rhythm matching rate deviation index is:

[0098] Collect the rhythm data of the music searched by the user, mainly including the beat rate, and at the same time collect the BPM values of each music in the recommended music list; perform standardization processing on the BPM values to generate two standardized BPM sequences, one corresponding to the music searched by the user and the other corresponding to the recommended music; for each music searched by the user and each music in the recommended list, calculate their rhythm matching rate , and the expression is: ; In the formula, is the BPM value of the music searched by the user, is the BPM value of the recommended music. Average the RMR values of all music pairs to obtain the overall rhythm matching rate , and the expression is: ; In the formula, n is the number of pairs of music searched by the user and recommended music. Calculate the rhythm matching deviation for each pair of music, and the expression is: ;Average the matching deviations of all music pairs to obtain the rhythm matching rate deviation index RMRD, and the expression is: 。

[0099] The larger the rhythm matching rate deviation index, the greater the deviation in the rhythm matching between the recommended music and the music searched by the user, and the poorer the consistency of the recommended content in this key dimension. This indicates that the global music preference model fails to accurately capture the user's preference for rhythm, resulting in the recommended results deviating from the user's actual needs. Therefore, the larger the rhythm matching rate deviation index, usually the lower the accuracy of the global music preference model, especially in the understanding and application of rhythm features, which need to be optimized and improved.

[0100] Convert the track jump density fluctuation index and the rhythm matching rate deviation index into the first feature vector, and use the first feature vector as the input of the machine learning model. The machine learning model takes predicting the accuracy value label of the global music preference model for each group of the first feature vectors as the prediction target, and minimizing the sum of the prediction errors of the accuracy value labels of all global music preference models as the training target, and trains the machine learning model until the sum of the prediction errors reaches convergence and then stops the model training. Determine the accuracy value of the global music preference model according to the model output result, where the machine learning model is a polynomial regression model.

[0101] The method for obtaining the accuracy value of the global music preference model is: obtain the corresponding function expression from the first feature vector training data of the trained machine learning model: ;In the formula, is the output function of the model, is the rhythm matching rate deviation index, TSD is the track jump density fluctuation index, is the accuracy value of the global music preference model.

[0102] S4: According to the evaluation results, divide the accuracy of the global music preference model into different levels, divide it into high accuracy level, general accuracy level and low accuracy level, and make corresponding adjustments to the music play recommendation list.

[0103] Compare the obtained accuracy value of the global music preference model with the gradient standard thresholds. The gradient standard thresholds include the first standard threshold and the second standard threshold, and the first standard threshold is less than the second standard threshold. Compare the accuracy value of the global music preference model with the first standard threshold and the second standard threshold respectively;

[0104] If the accuracy value of the global music preference model is greater than the second standard threshold, it means that the accuracy of the global music preference model is high, divide it into the high accuracy level, and generate a model high accuracy signal;

[0105] If the accuracy value of the global music preference model is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it indicates that the accuracy of the global music preference model is average. It is classified into the medium accuracy level, and a model medium accuracy signal is generated.

[0106] If the accuracy value of the global music preference model is less than the first standard threshold, it indicates that the accuracy of the global music preference model is low. It is classified into the low accuracy level, and a model low accuracy signal is generated.

[0107] Adjustment for the high accuracy level: Since the accuracy of the model is already very high, the recommendation strategy generally does not need to be significantly adjusted. Continue to use the existing model to generate the recommendation list. On the basis of maintaining the main recommended content, add a certain proportion of novel or untried music to expand the user's interest range. For example, add some tracks that are slightly different from the current preferences at the end of the recommendation list to encourage users to explore. Further refine the personalized recommendation by considering more detailed user behavior data (such as the user's listening time period, mood tags, environmental background, etc.) to optimize the recommended content. Make the recommendation more in line with the user's immediate needs and scenarios, and enhance the accuracy of the recommendation.

[0108] Adjustment for the medium accuracy level: Analyze the features with higher and lower accuracy in the user preference model, and identify the features that have a greater impact on the recommendation effect for optimization. Increase the weight of important user features (such as specific music genres or artists), while reducing the influence of irrelevant features. Increase the proportion of the content that the user is most interested in the recommendation list to enhance the overall recommendation effect. Introduce a hybrid recommendation strategy on the basis of the existing model, combining collaborative filtering, content-based recommendation, and context-based recommendation to provide diversified recommended content. For example, use collaborative filtering to recommend the music liked by other users with a high similarity to the user, or recommend relevant music based on the user's current situation (such as weather, geographical location). Through diverse recommendation methods, improve the user's acceptance of the recommended content. Strengthen the real-time collection and analysis of user feedback, and dynamically adjust the recommendation list according to the user's click, skip, favorite and other behaviors. According to the user's actual operations on the recommendation list, adjust the recommended content in a timely manner. For example, reduce the track types with a high user skip rate and increase the track types with a high user favorite rate. Continuously optimize the recommended content through user feedback, and gradually improve the accuracy of the model.

[0109] Adjustments at low accuracy levels: At low accuracy levels, the model may need to be retrained or completely updated. Introduce new data or algorithms to optimize model parameters. Introduce more user data (such as longer historical behavior data, more behavior features), or try different algorithms (such as deep learning, reinforcement learning) to improve model performance. Significantly improve the accuracy of recommendations through retraining or model updates. During the model optimization process, temporarily introduce popular recommendations or chart recommendations to reduce reliance on personalized recommendations. Use global popular charts, regional charts, or recent popular tracks to fill the recommendation list to maintain user interest and engagement. When personalized recommendations are not effective, use popular recommendations to maintain user experience. Re-establish the user's music preference model by collecting more user behavior data or through questionnaires. For example, show users a set of music clips and ask them about their preferences for these songs to obtain more accurate preference data. Improve the accuracy of the model through more accurate user preference reconstruction. Conduct A / B tests or multi-armed bandit experiments on different recommendation strategies to find the best recommendation strategy. Try different types of music or recommendation methods in the recommendation list, observe user reactions, and adjust the recommendation algorithm based on the experimental results. Explore more effective recommendation strategies and improve model performance.

[0110] S5: Based on the adjusted music play recommendation list, the user's music preference data is shared and migrated between different music platforms, and real-time monitoring is performed to obtain the data synchronization efficiency fluctuation status and the stability of the communication network during data transmission.

[0111] Obtain user authorization: Users first need to authorize on each music platform to allow the system to access and migrate their music preference data. OAuth 2.0 or other authentication protocols are used for user identity authentication and permission acquisition.

[0112] Multi-platform authentication: Ensure that users are authenticated on all participating music platforms and allow data access and transfer. Ensure the legality and security of the data sharing and migration process to prevent unauthorized operations.

[0113] Music preference data collection: Collect users' music preference data from various music platforms, including playback history, search history, favorites list, likes and other user behavior data.

[0114] Standardize the data formats of different platforms to ensure consistency in data format and structure so that they can be processed uniformly during the migration process.

[0115] Data cleaning: remove redundant data, correct data format errors, and ensure data accuracy and consistency. Ensure the accuracy and consistency of data before migration to provide a reliable data foundation for subsequent sharing and migration.

[0116] Data Encryption: Encrypt the user's music preference data using a strong encryption algorithm (such as AES-256) to ensure the security of the data during transmission and prevent data leakage or tampering.

[0117] Data Packing: Pack the encrypted data into a standardized format (such as JSON, XML, or an encrypted compressed package) for easy transmission and parsing between different platforms. Protect the privacy of user data through encryption and pack the data for efficient transmission.

[0118] Select Transmission Protocol: Use a secure transmission protocol (such as HTTPS, SFTP) to transfer the encrypted and packed music preference data from the source platform to the target platform.

[0119] Parallel Transmission and Incremental Synchronization: Transmit the data in parallel to optimize the transmission speed and efficiency. Only transmit the incremental data since the last synchronization to reduce the amount of data transmitted.

[0120] Transmission Monitoring and Verification: Monitor the transmission process in real-time to ensure the integrity and security of the transmitted data. Conduct integrity verification (such as using SHA-256 hash verification) after the transmission is complete to ensure that the data has not been tampered with. Ensure the secure and efficient transmission of data between different platforms and prevent data loss or damage.

[0121] Data Decryption: After the target platform receives the encrypted data, use the corresponding decryption key to decrypt the data and restore it to usable plaintext data.

[0122] Data Parsing and Integration: Parse the decrypted data and convert it into a format recognizable by the target platform. Integrate the migrated data with the existing user data on the target platform to ensure data consistency and integrity.

[0123] Data Conflict Handling: If the target platform already has the same data, conflict handling is required to determine whether to overwrite, merge, or retain multiple versions of the data. Ensure the effective utilization of user data on the target platform and maintain compatibility with the existing data.

[0124] Real-time monitor the state of data synchronization efficiency fluctuations during the data transmission process. Use logging, bandwidth monitoring, and error detection tools to continuously track the changes in these metrics. By analyzing their fluctuation trends, timely identify potential network bottlenecks or transmission failures. If abnormal fluctuations are detected, the system can automatically trigger an alarm and take corresponding adjustment measures to ensure the stability and efficiency of data transmission.

[0125] Generate a transmission delay jitter index based on the state of data synchronization efficiency fluctuations during the data transmission process. The method for obtaining the transmission delay jitter index is as follows:

[0126] Collect the delay time series of each data packet during the data transmission process, and perform standardization or normalization on the delay data so that the data can be analyzed on the same scale. The standardized delay time series , where n is the number of the data packet.

[0127] Select a wavelet function ψ(s) to decompose the delay data, where s is time. Commonly used wavelet functions include Haar wavelet, Daubechies wavelet (such as db4), etc. The specific selection depends on the characteristics of the signal and the purpose of analysis. Determine the number of decomposition levels J of the wavelet, usually selected according to the length and frequency characteristics of the delay signal. A higher number of decomposition levels can capture finer details. The selected wavelet function ψ(t) and the number of decomposition levels J.

[0128] Use the discrete wavelet transform to decompose the delay signal into approximation coefficients and detail coefficients at different scales. The expression is: ; is the coefficient at the j-th scale, is the corresponding wavelet basis function. Extract the detail coefficients at each level , and the expression is: ; These coefficients reflect the rapidly changing part of the delay signal at different frequencies. The detail coefficients at each level , , ... , where represents the detail coefficient at the j-th level. Calculate the energy of the detail coefficient at each level , and the expression is: ; The energy reflects the magnitude of the coefficients at this level, that is, the degree of signal fluctuation. Among them, is the k-th detail coefficient at the j-th level, is the corresponding energy. The energies of the detail coefficients at each level are , , ... ; Combine the energies of the detail coefficients at all levels to obtain the total jitter energy , and the expression is: ; Perform normalization on the total energy to obtain the transmission delay jitter index , and the expression is: ; where N is the total length of the delay signal.

[0129] The larger the transmission delay jitter index, it indicates that during the music preference data sharing process, the transmission delay of data packets fluctuates greatly, and the stability of network transmission is poor. Such fluctuations may increase the risk of data packet loss, error retransmission, or transmission interruption, thereby reducing the reliability and security of data transmission. Therefore, the larger the transmission delay jitter index, usually it means the lower the security of the music preference data sharing process, and it may be necessary to strengthen the stability of network transmission and data integrity guarantee measures.

[0130] Generate a retransmission aggregation index according to the stability of the communication network during the data transmission process, and the method for obtaining the retransmission aggregation index is as follows:

[0131] Cluster the retransmission events according to the time ti and network location pi, and divide the retransmission events into several clustering clusters C1, C2, …, Cm, where m is the number of clustering clusters, and count the number of retransmission events in each clustering cluster , where j = 1, 2, …, m; the retransmission event clusters after clustering and the number of retransmissions within each cluster.

[0132] Calculate the weight of each clustering cluster , that is, the proportion of the number of retransmissions in each cluster to the total number of retransmissions, and the expression is: ; in the formula, n is the total number of retransmission events, calculate the sum of the squares of the weights of all clusters to represent the aggregation degree, that is, calculate the retransmission aggregation index, and the expression is: ; in the formula, is the retransmission aggregation index.

[0133] In this application, the retransmission aggregation index quantifies the aggregation of retransmission events during the music data transmission process through clustering analysis and weight calculation. A higher retransmission aggregation index indicates that retransmission events occur concentratedly in certain time periods or network paths, which may indicate local problems in the network. By monitoring and analyzing the retransmission aggregation index, it can help identify network bottlenecks, optimize data transmission strategies, and improve the stability and reliability of music data sharing.

[0134] S6: Analyze the data synchronization efficiency fluctuation state and the stability of the communication network during the data transmission process through fuzzy logic, evaluate the security of the music preference data sharing process, and optimize the music preference data sharing process according to the evaluation results.

[0135] Take the transmission delay jitter index TD and the retransmission aggregation index MR as the input items of fuzzy logic, and take the security value WD of the music preference data sharing process as the output item of fuzzy logic;

[0136] Define fuzzy sets for each input variable TD and MR, such as "Low", "Medium", "High".

[0137] Define similar fuzzy sets for the output variable WD, such as "Low Safety", "Medium Safety", and "High Safety".

[0138] Design membership functions: Use triangular or trapezoidal membership functions to define the degree of membership for each fuzzy set. That is, define membership functions for the retransmission aggregation degree index MR and the output safety value WD.

[0139] Fuzzification process: For specific TD and MR values, calculate their degrees of membership in each fuzzy set. For example, if TD = 0.5 and the membership function is μLow(0.5) = 0.7, then the degree of membership of TD in the "Low" set is 0.7.

[0140] The fuzzy membership degrees of the input variables TD and MR serve as the basis for fuzzy logic rules.

[0141] Based on experience and domain knowledge, design fuzzy rules to map the input variables to the output variable. For example: If TD is "High" and MR is "High", then WD is "Low Safety". If TD is "Low" and MR is "Low", then WD is "High Safety". Other combinations can be deduced by analogy.

[0142] Construct a fuzzy rule table to map all possible combinations of TD and MR to the corresponding WD output.

[0143] According to the current TD and MR values, activate the corresponding fuzzy rules. The activated rules are determined by the input membership degrees.

[0144] For example, if the degree of membership of TD in "Medium" is 0.6 and in "High" is 0.4, and at the same time, the degree of membership of MR in "Medium" is 0.7 and in "High" is 0.3, then the activation strength of the corresponding rule such as "TD = Medium and MR = Medium" is 0.42 (0.6 × 0.7).

[0145] Perform inference on all activated rules to calculate the degrees of membership of the output variable WD in each fuzzy set.

[0146] Use the min - max inference method or the fuzzy weighted average method to synthesize the outputs of the activated rules. The expression is: ; where and are the input membership functions, is the output membership function;

[0147] Use the Centroid Method to defuzzify the output membership function to obtain the precise safety value WD. The expression is: ; where y is a possible security value, is the corresponding membership degree.

[0148] Based on the calculated security value, judge the security of the current data sharing process, compare the obtained security value with the security value reference threshold. If the security value is greater than or equal to the security value reference threshold, it indicates a lower risk. At this time, no warning signal is generated and no processing is required; if the security value is less than the security value reference threshold, it indicates a greater risk. At this time, a warning signal is generated and optimization measures need to be taken.

[0149] The optimization strategy can be: by adjusting the network configuration, reducing the transmission delay jitter index TD, such as optimizing the routing and increasing the bandwidth. By improving the transmission protocol or enhancing the channel quality, reducing the retransmission aggregation index MR, such as using error detection and correction techniques. Establish a dynamic monitoring system, adjust the transmission strategy according to real-time feedback, and ensure the continuous stability of the transmission.

[0150] In this embodiment, by integrating the music preference data of users on multiple music platforms, constructing a global music preference model and optimizing the recommendation list in real time, it is ensured that the recommended content highly matches the user's interests. Using fuzzy logic to analyze the synchronization efficiency and network stability in the data transmission process can evaluate and improve the security and reliability of the music data sharing process. This systematic method not only enhances the accuracy of recommendations and personalized experience, but also ensures that the transmission stability and security of user data are guaranteed during the data sharing process between multiple platforms, thereby improving the overall user experience and system performance.

[0151] Embodiment 2, please refer to Figure 2 As shown, a music intelligent recommendation and music data sharing system in this embodiment includes a data acquisition module, a global music preference model construction module, an accuracy evaluation module, an adjustment module, a data sharing module, and a security analysis module;

[0152] Data acquisition module: Acquire the music preference data of users on several music platforms, and the music preference data includes music play data, search data, and user interaction data;

[0153] Global music preference model construction module: According to the music preference data of users on each platform, construct a global music preference model of the user, and generate a real-time music play recommendation list through the global music preference model;

[0154] Accuracy evaluation module: Evaluate the accuracy of the global music preference model according to the change of the play rate of the music play recommendation list by the user and the similarity degree between the music searched by the user and the music recommended in the music play recommendation list;

[0155] Adjustment module: According to the evaluation results, divide the accuracy of the global music preference model into different levels, divide it into high-accuracy level, general-accuracy level and low-accuracy level, and make corresponding adjustments to the music play recommendation list;

[0156] Data sharing module: According to the adjusted music play recommendation list, share and migrate the music preference data of the user between different music platforms, and monitor in real time to obtain the data synchronization efficiency fluctuation state during the data transmission process and the stability of the communication network;

[0157] Security analysis module: Analyze the data synchronization efficiency fluctuation state during the data transmission process and the stability of the communication network through fuzzy logic, evaluate the security of the music preference data sharing process, and optimize the music preference data sharing process according to the evaluation results.

[0158] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to get a formula closest to the real situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0159] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center in a wired or wireless (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0160] It should be understood that the term "and / or" in this text is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. Additionally, the character " / " in this text generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood by referring to the context before and after.

[0161] It should be understood that in various embodiments of the present application, the magnitudes of the serial numbers of the above processes do not indicate the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0162] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this text can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0163] As described above, this is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application.

Claims

1. A method for intelligent music recommendation and music data sharing, characterized in that: The following steps are involved: S1: Obtaining music preference data of users on several music platforms, wherein the music preference data includes music playing data, search data and user interaction data; S2: Build a global music preference model for users based on their music preference data on various platforms, and generate a real-time music play recommendation list through the global music preference model; S3: Evaluate the accuracy of the global music preference model based on the changes in the user's play rate of the music play recommendation list and the similarity between the user's searched music and the music recommended by the music play recommendation list; S4: According to the evaluation results, the accuracy of the global music preference model is divided into different levels, namely, a high accuracy level, a general accuracy level and a low accuracy level, and the music playback recommendation list is adjusted accordingly; S5: Based on the adjusted music play recommendation list, the user's music preference data is shared and migrated between different music platforms, and real-time monitoring is performed to obtain the fluctuation status of data synchronization efficiency and the stability of the communication network during data transmission; S6: Analyze the fluctuation state of data synchronization efficiency and the stability of the communication network during data transmission through fuzzy logic, evaluate the security of the music preference data sharing process, and optimize the music preference data sharing process based on the evaluation results.

2. The method for intelligent music recommendation and music data sharing according to claim 1, characterized in that: In S3, the accuracy of the global music preference model is evaluated based on the change in the user's play rate of the music play recommendation list and the similarity between the user's searched music and the music recommended by the music play recommendation list, specifically: The track jump density fluctuation index is generated according to the change of the play rate of the user's music play recommendation list. The method for obtaining the track jump density fluctuation index is: The track jump density of the user in a time window is calculated and recorded as the time series y1, y2, …, yt, …, yT, where yt represents the track jump density at time t, and the jump behavior is divided into three states: low jump state (L), medium jump state (M), and high jump state (H); the time series yt is discretized into the observation sequence O={o1, o2, …, oT}, where ot represents the discrete observation value at time t; the initial state probability vector π={π1, π2, …, πN}: represents the probability that the system belongs to state i at the initial moment, where N is the number of states, and the state transition matrix A={aij}: represents the probability of transferring from state i to state j, where ;Observation probability matrix B={bj(o)}: represents the probability of observing ot=o in state j; Use the Baum-Welch algorithm to train it, update the model parameters λ=(π,A,B), adjust the parameters by maximizing the likelihood function of the observation sequence, and iterate repeatedly on the given training data until the parameters converge; use the Viterbi algorithm to find the target state sequence S={s1,s2,…,sT}, that is, the user's jump behavior pattern in different time periods, expressed as: ; In the formula, is the optimal path probability in the Viterbi algorithm, Indicates at time The probability of the optimal path of the system in state i at the moment is obtained by tracing the optimal state sequence backward from the final time T, and calculating the transition frequency between the high jump density state (H) and the low jump density state (L), that is, calculating the track jump density fluctuation index, which is expressed as: ; Where TSD is the track jump density fluctuation index.

3. A method for intelligent music recommendation and music data sharing according to claim 2, characterized in that: In S3, a rhythm matching rate deviation index is generated according to the similarity between the music searched by the user and the music recommended in the music play recommendation list. The rhythm matching rate deviation index is obtained as follows: Collect the rhythm data of the music that users search for, including the beat rate, and collect the BPM value of each piece of music in the recommended music list; standardize the BPM value to generate two standardized BPM sequences, one for the music that users search for and the other for the recommended music; calculate the rhythm matching rate for each piece of music that users search for and each piece of music in the recommended list , the expression is: ; In the formula, is the BPM value of the music that the user searches for, is the BPM value of the recommended music. The RMR values ​​of all music pairs are averaged to get the overall rhythm matching rate. , the expression is: ; In the formula, n is the number of user search music and recommended music pairs, and the rhythm matching deviation of each pair of music is calculated , the expression is: ; The matching deviations of all music pairs are averaged to obtain the rhythm matching rate deviation index RMRD, which is expressed as: .

4. The method for intelligent music recommendation and music data sharing according to claim 3, characterized in that: In S3, the accuracy of the global music preference model is evaluated, specifically: The track jump density fluctuation index and the rhythm matching rate deviation index are converted into the first eigenvector, and the first eigenvector is used as the input of the machine learning model. The machine learning model uses each group of first eigenvectors to predict the accuracy value label of the global music preference model as the prediction target, and takes minimizing the sum of prediction errors of the accuracy value labels of all global music preference models as the training target. The machine learning model is trained until the sum of prediction errors converges, and the model training is stopped. The accuracy value of the global music preference model is determined according to the model output results, wherein the machine learning model is a polynomial regression model.

5. A method for intelligent music recommendation and music data sharing according to claim 4, characterized in that: In S4, the accuracy of the global music preference model is divided into different levels, specifically: Comparing the acquired accuracy value of the global music preference model with a gradient standard threshold, the gradient standard threshold includes a first standard threshold and a second standard threshold, and the first standard threshold is less than the second standard threshold, and comparing the accuracy value of the global music preference model with the first standard threshold and the second standard threshold respectively; If the accuracy value of the global music preference model is greater than the second standard threshold, it indicates that the accuracy of the global music preference model is high, and the model is classified as a high accuracy level, and a model high accuracy signal is generated; If the accuracy value of the global music preference model is greater than or equal to the first standard threshold and less than or equal to the second standard threshold, it means that the accuracy of the global music preference model is average, and the model is classified as a medium accuracy level, and a model medium accuracy signal is generated; If the accuracy value of the global music preference model is less than the first standard threshold, it indicates that the accuracy of the global music preference model is low, and the model is classified as a low accuracy level, and a model low accuracy signal is generated.

6. The method for intelligent music recommendation and music data sharing according to claim 1, characterized in that: In S5, real-time monitoring is performed to obtain the fluctuation state of data synchronization efficiency and the stability of the communication network during data transmission, specifically: The transmission delay jitter index is generated according to the data synchronization efficiency fluctuation state during the data transmission process. The method for obtaining the transmission delay jitter index is: Collect the delay time series of each data packet during data transmission, standardize the delay data, and the standardized delay time series , where n is the number of the data packet; select the wavelet function ψ(s) to decompose the delayed data, s is the time, determine the number of layers J of the wavelet decomposition, and use discrete wavelet transform to transform the delayed signal Decomposed into approximation coefficients and detail coefficients at different scales, the expression is: ; is the coefficient at the jth level scale, is the corresponding wavelet basis function, extracting the detail coefficients of each layer , the expression is: Detail coefficient of each layer , , ... ,in, Represents the detail coefficient of the jth layer, and calculates the energy of the detail coefficient of each layer , the expression is: ;in, is the kth detail coefficient of the jth layer, is the corresponding energy, and the energy of the detail coefficients of each layer is , , ... ; Combine the detail coefficient energies of all layers to get the total jitter energy , the expression is: ; Normalize the total energy to get the transmission delay jitter index , the expression is: ; where N is the total length of the delayed signal.

7. A method for intelligent music recommendation and music data sharing according to claim 6, characterized in that: The retransmission concentration index is generated according to the stability of the communication network during data transmission. The method for obtaining the retransmission concentration index is: Cluster the retransmission events according to time ti and network position pi, and divide the retransmission events into several clusters C1, C2, ..., Cm, where m is the number of clusters, and count the number of retransmission events in each cluster , where j=1,2,…,m; clustered retransmission event clusters and the number of retransmissions in each cluster; calculate the weight of each cluster , that is, the ratio of the number of retransmissions of each cluster to the total number of retransmissions, expressed as: ; Where n is the total number of retransmission events, and the weighted sum of all clusters is calculated to represent the aggregation, that is, the retransmission aggregation index is calculated, and the expression is: ; In the formula, is the retransmission concentration index.

8. The method for intelligent music recommendation and music data sharing according to claim 7, characterized in that: In S6, the data synchronization efficiency fluctuation state and the stability of the communication network during data transmission are analyzed through fuzzy logic to evaluate the security of the music preference data sharing process. Specifically: The transmission delay jitter index TD and the retransmission concentration index MR are used as the input items of fuzzy logic, and the security value WD of the music preference data sharing process is used as the output item of fuzzy logic; Define fuzzy sets for each input variable TD and MR, and also define fuzzy sets for the output variable WD; Define the membership degree for each fuzzy set, that is, define the membership function for the retransmission aggregation index MR and the output security value WD; For specific TD and MR values, calculate their membership in each fuzzy set; The fuzzy membership of input variables TD and MR is used as the basis of fuzzy logic rules; Design fuzzy rules to map input variables to output variables; Construct a fuzzy rule table to map all TD and MR combinations to corresponding WD outputs; According to the current TD and MR values, the corresponding fuzzy rules are activated; Reasoning is performed on all activated rules, and the membership of the output variable WD in each fuzzy set is calculated; the output of the activated rules is integrated using the minimum-maximum reasoning method, and the expression is: ; In the formula, and is the input membership function, is the output membership function; The output membership function is defuzzified using the centroid method to obtain the precise safety value WD, which is expressed as: ;in, is the corresponding degree of membership.

9. A method for intelligent music recommendation and music data sharing according to claim 8, characterized in that: According to the calculated security value, the security of the current data sharing process is judged, and the obtained security value is compared with the security value reference threshold. If the security value is greater than or equal to the security value reference threshold, no warning signal is generated and no processing is required; If the safety value is less than the safety value reference threshold, a warning signal is generated and optimization measures are taken.

10. A music intelligent recommendation and music data sharing system, used to implement a music intelligent recommendation and music data sharing method according to any one of claims 1 to 9, characterized in that: It includes data acquisition module, global music preference model building module, accuracy assessment module, adjustment module, data sharing module and security analysis module; Data acquisition module: acquires music preference data of users on several music platforms, wherein the music preference data includes music playing data, search data and user interaction data; Global music preference model building module: Build a global music preference model for users based on their music preference data on various platforms, and generate a real-time music play recommendation list through the global music preference model; Accuracy evaluation module: evaluates the accuracy of the global music preference model based on the changes in the user's play rate of the music play recommendation list and the similarity between the user's searched music and the music recommended by the music play recommendation list; Adjustment module: According to the evaluation results, the accuracy of the global music preference model is divided into different levels, namely, high accuracy level, general accuracy level and low accuracy level, and the music playback recommendation list is adjusted accordingly; Data sharing module: Based on the adjusted music play recommendation list, the user's music preference data is shared and migrated between different music platforms, and real-time monitoring is performed to obtain the fluctuation status of data synchronization efficiency and the stability of the communication network during data transmission; Security analysis module: Analyze the fluctuation of data synchronization efficiency and the stability of the communication network during data transmission through fuzzy logic, evaluate the security of the music preference data sharing process, and optimize the music preference data sharing process based on the evaluation results.

Images