User music recommendation method and system based on data analysis

Abstract

The invention relates to the technical field of content recommendation, in particular to a user music recommendation method and system based on data analysis, and the method comprises the following steps: based on a music playing record of a user, extracting the playing frequency and frequency information of each piece of music, evaluating the preference degree of the user for each music type, analyzing the preference degree of the user for each music label, and recommending the user music. And obtaining a user preference analysis result. According to the method, the preference degrees of the users for different music types and labels are analyzed according to the playing frequency and frequency, the ages and genders of the users are considered, the fineness and adaptability of recommendation are optimized, user groups with similar music preferences can be identified more accurately through evaluation and classification of the similarity between the users, and the user experience is improved. A more accurate recommendation list is provided, the playing habits of the user in different time periods are analyzed, the recommendation opportunity is optimized, it is ensured that the user receives the recommendation in an acceptable time period, the participation degree of the user is improved, and the acceptance degree of the user to the recommended content is increased.

Description

Technical Field

[0001] The present invention relates to the technical field of content recommendation, and in particular to a method and system for user music recommendation based on data analysis. Background Art

[0002] The field of content recommendation technology involves the use of algorithms and data analysis methods to push personalized content to users. The core purpose of the technology is to predict the content that users may like and be interested in by analyzing their historical behavior, preference settings, contextual environment and other relevant data. Commonly used technologies include collaborative filtering, content-based filtering, hybrid recommendation systems, etc. These systems are able to process large amounts of dynamic data and update recommendation results in real time, thereby increasing user interaction and satisfaction. Content recommendation technology is widely used in e-commerce, social networks, online video and music services, and is a key tool for improving user experience and business growth.

[0003] Among them, the user music recommendation method of data analysis involves analyzing user data and recommending music content for them. The method identifies the user's music preferences based on the user's listening history, search habits, rating feedback, and social media activities, and recommends new songs or artists to the user to enhance user experience and engagement. The purpose of the method is to improve user satisfaction and retention, and to enhance user interaction and brand loyalty by providing a customized music experience. The recommendation method can also help music platforms position themselves in the market more effectively and promote the discovery and consumption of related music.

[0004] Although existing content recommendation technologies can provide personalized recommendations based on user historical behaviors, common methods such as collaborative filtering and content-based filtering often ignore the real-time changes and diverse needs of users. These technologies may not be flexible enough when dealing with dynamically changing user preferences, resulting in the inability of recommended content to reflect the user's latest interests in a timely manner. In addition, when considering individual differences among users, existing technologies do not analyze socio-demographic characteristics such as age and gender in depth, which may lead to a lack of personalization in recommendation results in some cases. For example, young and old users may have very different preferences for music, but if the recommendation system simply makes recommendations based on historical playback data, it may not be able to effectively distinguish these differences, thus affecting the accuracy of recommendations and user satisfaction. These deficiencies limit the improvement of user experience and the potential for business growth, and sometimes even lead to user churn. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a user music recommendation method and system based on data analysis.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a user music recommendation method based on data analysis, comprising the following steps: S1: Based on the user's music playing records, extract the playing frequency and number of times of each piece of music, evaluate the user's preference for each type of music, analyze the user's preference for each music tag, and obtain the user preference analysis results; S2: Based on the user preference analysis results, extract each user's preference data for differentiated music types and music tags, extract the user's age and gender data, evaluate the similarity between users, classify the users, and obtain user classification results; S3: Based on the user classification results, according to the number of times and frequency of playing each piece of music by users in the same category, evaluate the popularity of each piece of music among users in the current category, sort the popularity of the music, and obtain a popular music list; S4: Based on the popular music list, and by analyzing the number of times the user plays music in differentiated time periods, identify the user's preferred music playing time period, extract the music that has not been played in the popular music list, make music recommendations to the user, and analyze the number of times the user plays the recommended music, frequency and sharing data, evaluate the user's acceptance of the recommended music, and obtain music recommendation results.

[0007] The present invention is improved in that the step of evaluating the user's preference for each type of music is as follows: S111: Based on the user's music playing records, extract the playing frequency and number of times of each piece of music, and extract the type information of each piece of music, count the number of times each type of music is played and the daily playing frequency within a period of time, and obtain type preference association data; S112: Based on the type preference association data, by formula: ; Calculate the user's preference index for music types and obtain the user's music type preference evaluation result; in, is the total number of times a music type is played. is the total number of times all types of music are played. is the average daily frequency of music genre playback, and is the weight parameter, It is the user's preference index for music types.

[0008] The present invention is improved in that the steps of obtaining the user preference analysis result are: S121: extracting the playing frequency and number of times of each piece of music based on the music playing record of the user, counting the number of times the tags of the music played by the user in a period of time, and obtaining music tag associated data; S122: Based on the music tag association data, by formula: ; Calculate the user's preference index for music tags; in, and is the weight parameter, is the exponential parameter, It's a music label The number of occurrences during the observation period, is the total number of occurrences of all music tags during the observation period, Is the user's music tag The preference index of S123: Based on the user's preference index for music tags, the user's music type preference evaluation results are integrated to obtain a user preference analysis result.

[0009] The present invention is improved in that the steps of obtaining the user classification results are: S211: Based on the user preference analysis result, extract each user's preference data for differentiated music types and music tags, and extract the user's age and gender data to obtain user category association data; S212: Based on the user category association data, by formula: ; Calculate the similarity between users; in, is the similarity between user 1 and user 2, and Represent the ages of user 1 and user 2 respectively. and Represents the gender of user 1 and user 2 respectively, is the gender comparison function. If the users have the same gender, , then it returns 1, otherwise it returns 0. is the weight of age difference, For equal weights for both genders, and Represents the responses of user 1 and user 2 to The preference index of the music genre, and Represents the music tags of user 1 and user 2 respectively The preference index, For the The weight of the music genre, For music labels The weight of is the total number of music types, is the total number of music tags, is the base of natural logarithms; S213: Based on the similarity between the users, the similarity is compared with a preset similarity threshold, and the users whose similarity is higher than the preset similarity threshold are marked as the same type of users, to obtain a user classification result.

[0010] The present invention is improved in that the steps of obtaining the popular music list are: S311: Based on the user classification result, extract the number of times and frequency of play of each piece of music by the current category of users to obtain music popularity association data; S312: Based on the music popularity association data, the formula: ; Calculate the popularity of each piece of music; in, For Music popularity, For Music Total number of plays among users in the current category, For Music The playback frequency, and is the weight coefficient; S313: Based on the popularity of each piece of music, the popularity of the music is sorted to obtain a popular music list.

[0011] The present invention is improved in that the step of identifying the music playing time period preferred by the user is as follows: S411: extracting the number of times the user plays music in different time periods based on the user's music playing records, and extracting the length data of the corresponding time periods to obtain playing time period related data; S412: Based on the play period association data, the formula: ; Calculate the user preference level in each time period; in, is the number of music plays in the time period, is the length of the period, is the weight parameter, is the user preference for the time period; S413: Based on the user preference level of each time period, by comparing the preference levels of the differentiated time periods, the time period with the greatest preference level is selected to obtain a target time period selection result.

[0012] The present invention is improved in that the steps of obtaining the music recommendation result are: S421: extracting the unplayed music in the popular music list based on the target time period selection result, recommending music to the user, and extracting the number of times, frequency and sharing data of the user for the recommended music to obtain music acceptance related data; S422: Based on the music, accept the associated data, by formula: ; Calculate the user's acceptance of the recommended music; in, is the user's acceptance of the recommended music. The number of times the user plays the recommended music. The frequency with which users play the recommended music. The number of times users share the recommended music. , and is the weight parameter; S423: Based on the user's acceptance level of the recommended music, the music is compared with a preset acceptance threshold. If the acceptance level is higher than the preset acceptance threshold, the user is marked as being able to receive the recommended music. Music recommendations are made based on the number of times the music of the same music producer is played to obtain music recommendation results.

[0013] A user music recommendation system for data analysis, the user music recommendation system for data analysis is used to execute the user music recommendation method for data analysis, the system comprising: The user preference evaluation module extracts the playing frequency and number of times of each piece of music based on the user's music playing records, evaluates the user's preference for each type of music, analyzes the user's preference for each music tag, and obtains the user preference analysis results; The user classification module extracts the preference data of each user for differentiated music types and music tags based on the user preference analysis results, extracts the user's age and gender data, evaluates the similarity between users, and classifies the users to obtain user classification results; The popular music identification module evaluates the popularity of each piece of music among users of the current category based on the user classification results and the number of times and frequency of each piece of music played by users of the same category, sorts the popularity of the music, and obtains a popular music list; The preferred time period analysis module is based on the user's music playing records and analyzes the number of times the user plays music in different time periods to identify the user's preferred music playing time period and obtain the target time period selection result; The music recommendation module extracts the unplayed music in the popular music list based on the target time period selection results and the popular music list, recommends music to the user, and analyzes the number of times, frequency and sharing data of the recommended music by the user, evaluates the user's acceptance of the recommended music, and obtains the music recommendation results.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by analyzing the user's preference for different music types and tags according to the frequency and number of times of playback, and taking the user's age and gender into consideration, the detailedness and adaptability of the recommendation are optimized. By evaluating and classifying the similarities between users, it is possible to more accurately identify user groups with similar music preferences, provide a more accurate recommendation list, analyze the user's playback habits in different time periods, optimize the recommendation timing, ensure that users receive recommendations in acceptable time periods, improve user participation, increase user acceptance of recommended content, improve user satisfaction and retention rate, and effectively support the market positioning of the music platform and the promotion of music consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flow chart of the method of the present invention; Figure 2 A flowchart for evaluating the user's preference for each type of music for the present invention; Figure 3 A flow chart of obtaining the user preference analysis result of the present invention; Figure 4 A flow chart of obtaining user classification results for the present invention; Figure 5 A flowchart for obtaining a popular music list for the present invention; Figure 6 A flowchart of the present invention for identifying a user's preferred music playing time period; Figure 7 The present invention is a flowchart for obtaining music recommendation results. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0017] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0018] See also Figure 1 The present invention provides a technical solution: a user music recommendation method based on data analysis, comprising the following steps: S1: Based on the user's music playing records, extract the playing frequency and number of times of each piece of music, evaluate the user's preference for each type of music, analyze the user's preference for each music tag, and obtain the user preference analysis results; S2: Based on the user preference analysis results, extract each user's preference data for differentiated music types and music tags, extract the user's age and gender data, evaluate the similarity between users, classify the users, and obtain the user classification results; S3: Based on the user classification results, the popularity of each piece of music among users in the current category is evaluated according to the number of times and frequency of each piece of music played by users in the same category, and the popularity of the music is sorted to obtain a popular music list; S4: Based on the popular music list, and by analyzing the number of times users play music in differentiated time periods, identify the user's preferred music playing time period, extract the music that has not been played in the popular music list, make music recommendations to users, and analyze the number of times users play the recommended music, frequency and sharing data, evaluate the user's acceptance of the recommended music, and obtain music recommendation results.

[0019] The results of user preference analysis include the preference index of each music type and the preference index of each music tag. The results of user classification include user group number, user attribute set and user preference matching degree. The popular music list includes the number of plays, play frequency and music popularity ranking of music in each category. The music recommendation results include information about recommended music, user feedback statistics and recommendation success rate data.

[0020] See also Figure 2 ,The steps to evaluate the user's preference for each type of music are: S111: Based on the user's music playing records, extract the playing frequency and number of times of each piece of music, and extract the type information of each piece of music, count the number of times each type of music is played and the daily playing frequency within a period of time, and obtain type preference association data; S112: Based on type preference association data, through the formula: ; Calculate the user's preference index for music types and obtain the user's music type preference evaluation result; in, is the total number of times a music type is played. is the total number of times all types of music are played. is the average daily frequency of music genre playback, and is the weight parameter, It is the user's preference index for music types.

[0021] formula: ; Parameter details and how to obtain them: : The total number of times a music type is played, which means the total number of times a certain music type is played within a certain period of time. The parameter is directly obtained from the music playing data. For example, if a user plays rock music 100 times in a week, then The value of is 100.

[0022] : The total number of times all types of music are played. This parameter is obtained by summing up the number of times the user plays all music in the time period. For example, if a user plays music 500 times in a week, then The value is 500, which indicates the user's overall attention to this music type during the observation period.

[0023] : The average daily frequency of music genres, that is, the average number of times a specific music genre is played each day. If a user plays rock music 10 times a day in a week, then is 10. By Divide by the number of days in the time period.

[0024] and : Weight parameter, used to adjust the influence of the proportion of play times and daily play frequency in user preference rating. Parameters are usually set based on historical data analysis and expert experience. For example, is 0.7, It is 0.3, which reflects that the playback proportion has a greater impact on preference than the daily frequency.

[0025] Calculation example: Set the following parameters: , , , , .

[0026] Calculate playback percentage: ; calculate : ; Calculation results , represents the user's preference score for the music type. In this example, due to the high influence of daily playback frequency, the user's preference for the music type is rated higher, indicating that the user has a high preference for this type of music. This calculation method can help the music recommendation system understand and predict the user's music preferences more accurately.

[0027] See also Figure 3 , the steps to obtain the user preference analysis results are: S121: extracting the playing frequency and number of times of each piece of music based on the music playing record of the user, counting the number of times the tags of the music played by the user in a period of time, and obtaining music tag associated data; S122: Based on the music tag association data, the formula is: ; Calculate the user's preference index for music tags; in, and is the weight parameter, is the exponential parameter, It's a music label The number of occurrences during the observation period, is the total number of occurrences of all music tags during the observation period, Is the user's music tag The preference index of S123: Based on the user's preference index for music tags, the user's music type preference evaluation results are integrated to obtain a user preference analysis result.

[0028] formula: ; Parameter details and how to obtain them: : Music Tags The number of occurrences during the observation period is obtained by traversing the user's music playing records and counting the frequency of occurrence of the tags of each piece of music. For example, if the tag "sadness" appears 150 times in a week in the user's playlist, then for the tag "sadness", .

[0029] : The total number of occurrences of all music tags during the observation period is obtained by counting the total number of occurrences of all tags in the user's playlist. For example, if a user plays music with various tags 1,000 times in the same week, then .

[0030] , , : Weight coefficients and index parameters are usually set based on historical data analysis or expert recommendations. For example, if analysis of historical user data shows that the proportion of tags has a greater impact on user preferences, then it may be set , , To enhance the nonlinear effect of the proportion.

[0031] Calculation example: The setting data is as follows: , indicating the number of occurrences of the label “sadness”. , which represents the total number of occurrences of all tags. , , .

[0032] Calculate the label ratio: ; Calculate the preference index : ; Calculated , reflects the user's preference for the "sad" music tag. This value takes into account the relative frequency and total number of occurrences of the tag, and the calculation method of the preference can be adjusted according to the actual situation of different user groups. This method can help the music recommendation system understand the personalized needs of users more accurately, and then provide music recommendations that are more in line with their preferences.

[0033] See also Figure 4 , the steps to obtain the user classification results are: S211: based on the user preference analysis results, extracting each user's preference data for differentiated music types and music tags, and extracting the user's age and gender data to obtain user category association data; S212: Based on the user category association data, the formula is: ; Calculate the similarity between users; in, is the similarity between user 1 and user 2, and Represent the ages of user 1 and user 2 respectively. and Represents the gender of user 1 and user 2 respectively, is the gender comparison function. If the users have the same gender, , then it returns 1, otherwise it returns 0. is the weight of age difference, For equal weights for both genders, and Represents the responses of user 1 and user 2 to The preference index of the music genre, and Represents the music tags of user 1 and user 2 respectively The preference index, For the The weight of the music genre, For music labels The weight of is the total number of music types, is the total number of music tags, is the base of natural logarithms; S213: Based on the similarity between users, the similarity is compared with a preset similarity threshold, and users with similarity higher than the preset similarity threshold are marked as the same type of users to obtain a user classification result.

[0034] formula: ; Parameter details and how to obtain them: and : and It is the weight parameter of music type and music label in user similarity score. The weight is usually set based on the output of historical data analysis and machine learning model, reflecting the influence of different music types and labels on user preference similarity. For large music types and labels, the score can be smaller to reduce the influence of large music types and labels on similarity. For example, the music types are "pure music" and "orchestral music". Since "orchestral music" belongs to "pure music" in classification, the weight of "orchestral music" can be greater than that of "pure music", so as to achieve the effect of refining the similarity between users.

[0035] : is the weight of age difference, which is used to adjust the impact of age difference between users on their similarity score. This parameter is usually determined based on user age distribution and user behavior research to ensure that the impact of age difference on similarity is consistent with actual observation data.

[0036] : It is the bonus weight for the same gender, determined based on statistical analysis, and is used to enhance the role of gender similarity in user similarity scoring.

[0037] : Gender comparison function, if the users have the same gender ( ), returns 1, otherwise returns 0. This function is calculated directly based on the gender data and is a simple Boolean comparison.

[0038] Calculation example: Set the following parameter settings: only consider one music type and music label, and for the music type and label weights, set and . Age difference weight Gender similarity weight . Set the ages of user 1 and user 2 to 25 and 30 respectively, with an age difference of . Users have the same gender, .

[0039] Similarity score calculation for music genre and label: set up , ,but .

[0040] For the label, set , ,but .

[0041] Calculate the similarity contribution of music type and label: .

[0042] Effects of age differences: .

[0043] Age Difference Score: .

[0044] Gender similarity bonus: .

[0045] Similarity score: .

[0046] Calculated similarity score , indicating that the two users are relatively similar in terms of music preferences, age, and gender. Ratings can help the system to group and recommend users more effectively, taking into account user characteristics in multiple dimensions.

[0047] See also Figure 5 , the steps to obtain the popular music list are: S311: Based on the user classification results, extract the number of times and frequency of play of each piece of music by the current category of users to obtain music popularity association data; S312: Based on the music popularity correlation data, the formula is: ; Calculate the popularity of each piece of music; in, For Music popularity, For Music Total number of plays among users in the current category, For Music Frequency of playback among users in the current category, and is the weight coefficient; S313: Based on the popularity of each piece of music, the popularity of the music is sorted to obtain a popular music list.

[0048] formula: ; Parameter details and how to obtain them: :music The total number of plays among users in the current category. This parameter is obtained by aggregating all play records related to the user category in the database, ensuring that it accurately reflects the play frequency of the music in a specific user group.

[0049] :music The frequency of play among users in the current category, which indicates the number of plays per unit time. The data is usually calculated by analyzing user play logs within a specific time period and is used to measure the popularity of music and user activity.

[0050] and : Weight parameters that determine the relative importance of the number of plays and the frequency of play in the overall score. These parameters are adjusted based on historical data and pre-set strategies to reflect the actual impact of different music attributes on user preferences.

[0051] Calculation example: The following specific values ​​are set: : Indicates music Played 500 times. : Indicates music The playback frequency is 50 times per day. , : Weight parameter, where the number of plays has a higher weight.

[0052] Calculation process: Logarithmic part of the number of playbacks: ; Weight-adjusted play score: ; Play frequency root part: ; Weight-adjusted play frequency scores: ; Total score: ; Calculated popularity score Indicates that the given music The score combines the number of plays and the frequency of plays, ensuring fairness and accuracy of the score through weighted balancing. This method can help develop more effective music recommendation and marketing strategies to ensure that the content is highly matched with user preferences.

[0053] See also Figure 6 ,The steps of identifying the music playing time period preferred by the user are: S411: extracting the number of times the user plays music in different time periods based on the user's music playing records, and extracting the length data of the corresponding time periods to obtain playing time period related data; S412: Based on the play period association data, the formula: ; Calculate the user preference level in each time period; in, is the number of music plays in the time period, is the length of the period, is the weight parameter, is the user preference for the time period; S413: Based on the user preference level of each time period, by comparing the preference levels of the differentiated time periods, the time period with the greatest preference level is selected to obtain a target time period selection result.

[0054] formula: ; Parameter details and how to obtain them: : The number of times music is played in a certain period of time. This parameter is obtained by querying the database of user play records, and counts the number of times each user plays in a certain period of time, such as the number of times the music is played from 8 to 9 am every day.

[0055] : The length of the time period, in hours. The parameter is calculated directly from the start and end time of the play period. For example, the play period from 8 to 9 in the morning is 1 hour long.

[0056] : Weight parameter, used to adjust the influence of the logarithmic number of plays in the calculation. This parameter is usually set based on the results of historical data analysis, aiming to balance the impact of the number of plays in different time periods and ensure a more reasonable calculation result.

[0057] Calculation example: The number of times a user plays during a typical weekday morning period (e.g. 8am to 9am) , time period length Hours, weight parameters .

[0058] Calculate the logarithmic number of plays: ; calculate : ; The calculated results , reflects the user's preference for music playback during a specific time period (8 to 9 a.m.). This value is closer to 1, indicating that the user is relatively active during this time period and tends to play music at this time. This measurement method can help the recommendation system better understand the user's daily playback habits, thereby pushing more appropriate music content during the corresponding time period and improving user satisfaction and interaction rate.

[0059] See also Figure 7 , the steps to obtain music recommendation results are: S421: based on the target time period selection result, extract the unplayed music in the popular music list, recommend music to the user, and extract the number of times, frequency and sharing data of the user for the recommended music to obtain music acceptance related data; S422: Accepting associated data based on music, by formula: ; Calculate the user's acceptance of the recommended music; in, is the user's acceptance of the recommended music. The number of times the user plays the recommended music. The frequency with which users play the recommended music. The number of times users share the recommended music. , and is the weight parameter; S423: Based on the user's acceptance level of the recommended music, the music is compared with a preset acceptance threshold. If the acceptance level is higher than the preset acceptance threshold, the user is marked as being able to receive the recommended music. Music recommendations are made based on the number of times the music of the same music producer is played to obtain music recommendation results.

[0060] formula: ; Parameter details and how to obtain them: :The number of times the user plays the recommended music. The parameter is obtained by analyzing the user's play record database and directly calculating the total number of times a specific music is played in a specific time period.

[0061] : The frequency with which users play the recommended music, that is, the number of times it is played within a certain time window. This parameter is obtained by counting the number of play events within the time window, such as the number of times it is played within a day or a week.

[0062] : The number of times a user shares the recommended music. The parameter is obtained from the data interface of the social function, and counts the number of times a user shares the music within a specific period of time.

[0063] , and : Weight parameters are used to adjust the influence of the number of plays, play frequency, and number of shares in the final acceptance score. The parameters are set based on historical data analysis and understanding of the influence of different user behaviors.

[0064] Calculation example: Set the following parameters: the number of times a user plays a piece of recommended music , how often the music is played in a week , the number of times the user shared the music , the weight is set to , , .

[0065] Calculation of the logarithmic part of the number of playbacks: ; Calculation of the square root of the playback frequency: ; Calculation of the square of the number of shares: ; Calculation of comprehensive acceptance level: ; The calculated results , indicating that the user has a high degree of acceptance of this recommended music. Considering the influence of the number of plays, play frequency, and number of shares, this high score means that the user likes this song very much and may further promote it to others. This calculation method provides an effective way for the music recommendation system to evaluate the success of music and user satisfaction, thereby guiding future recommendation strategies.

[0066] A user music recommendation system based on data analysis, the user music recommendation system based on data analysis is used to execute the user music recommendation method based on data analysis, and the system comprises: The user preference evaluation module extracts the playing frequency and number of times of each piece of music based on the user's music playing records, evaluates the user's preference for each type of music, analyzes the user's preference for each music tag, and obtains the user preference analysis results; The user classification module extracts each user's preference data for differentiated music types and music tags based on the user preference analysis results, and extracts the user's age and gender data, evaluates the similarity between users, and classifies the users to obtain user classification results; The popular music identification module evaluates the popularity of each piece of music among users of the current category based on the user classification results and the number and frequency of each piece of music played by users of the same category, sorts the popularity of the music, and obtains a popular music list; The preferred time period analysis module is based on the user's music playing records and analyzes the number of times the user plays music in different time periods to identify the user's preferred music playing time period and obtain the target time period selection result; The music recommendation module selects the results and popular music list based on the target time period, extracts the unplayed music in the popular music list, recommends music to users, analyzes the number of times users play the recommended music, and shares the data, evaluates the user's acceptance of the recommended music, and obtains the music recommendation results.

[0067] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A user music recommendation method based on data analysis, characterized in that: The following steps are involved: Based on the user's music playing records, extract the playing frequency and number of times of each piece of music, evaluate the user's preference for each type of music, analyze the user's preference for each music tag, and obtain the user preference analysis results; Based on the user preference analysis results, extract each user's preference data for differentiated music types and music tags, extract the user's age and gender data, evaluate the similarity between users, classify the users, and obtain user classification results; Based on the user classification results, according to the number of times and frequency of playing each piece of music by users in the same category, the popularity of each piece of music among users in the current category is evaluated, and the popularity of the music is sorted to obtain a popular music list; Based on the popular music list, and by analyzing the number of times users play music in differentiated time periods, the user's preferred music playing time period is identified, and music that has not been played in the popular music list is extracted. Music recommendations are made to users, and the number of times users play the recommended music, frequency and sharing data are analyzed to evaluate the user's acceptance of the recommended music and obtain music recommendation results.

2. The user music recommendation method based on data analysis according to claim 1, characterized in that: The steps of evaluating the user's preference for each type of music are: Based on the user's music playing records, extract the playing frequency and number of times of each piece of music, and extract the type information of each piece of music, count the number of times each type of music is played and the daily playing frequency within a period of time, and obtain type preference association data; Based on the type of preference associated data, through the formula: ； Calculate the user's preference index for music types and obtain the user's music type preference evaluation result; in, is the total number of times a music type is played. is the total number of times all types of music are played. is the average daily frequency of music genre playback, and is the weight parameter, It is the user's preference index for music types.

3. The user music recommendation method based on data analysis according to claim 2, characterized in that: The steps for obtaining the user preference analysis result are: Based on the user's music playing records, extract the playing frequency and number of times of each piece of music, count the number of occurrences of the tags with which the user plays the music within a period of time, and obtain the music tag association data; Based on the music tag association data, the formula is: ； Calculate the user's preference index for music tags; in, and is the weight parameter, is the exponential parameter, It's a music label The number of occurrences during the observation period, is the total number of occurrences of all music tags during the observation period, Is the user's music tag The preference index of Based on the user's preference index for music tags, the user's music type preference evaluation results are integrated to obtain a user preference analysis result.

4. The user music recommendation method based on data analysis according to claim 1, characterized in that: The steps for obtaining the user classification result are: Based on the user preference analysis results, extract each user's preference data for differentiated music types and music tags, and extract the user's age and gender data to obtain user category association data; Based on the user category association data, by formula: ； Calculate the similarity between users; in, is the similarity between user 1 and user 2, and Represent the ages of user 1 and user 2 respectively, and Represents the gender of user 1 and user 2 respectively, is the gender comparison function. If the users have the same gender, , then it returns 1, otherwise it returns 0. is the weight of age difference, For equal weights for both genders, and Represents the responses of user 1 and user 2 to The preference index of the music genre, and Represents the music tags of user 1 and user 2 respectively The preference index, For the The weight of the music genre, For music labels The weight of is the total number of music types, is the total number of music tags, is the base of natural logarithms; Based on the similarity between the users, the similarity is compared with a preset similarity threshold, and users with similarity higher than the preset similarity threshold are marked as the same type of users to obtain a user classification result.

5. The user music recommendation method based on data analysis according to claim 1, characterized in that: The steps for obtaining the popular music list are: Based on the user classification results, extract the number of times and frequency of play of each piece of music by the current category of users to obtain music popularity correlation data; Based on the music popularity correlation data, the formula is: ； Calculate the popularity of each piece of music; in, For Music popularity, For Music Total number of plays among users in the current category, For Music Frequency of playback among users in the current category, and is the weight coefficient; Based on the popularity of each piece of music, the popularity of the music is sorted to obtain a popular music list.

6. The user music recommendation method based on data analysis according to claim 1, characterized in that: The step of identifying the music playing time period preferred by the user is as follows: Based on the user's music playing records, extract the number of times the user plays music in different time periods, and extract the length data of the corresponding time periods to obtain the playing time period related data; Based on the playback period associated data, by formula: ； Calculate the user preference level in each time period; in, is the number of music plays in the time period, is the length of the period, is the weight parameter, is the user preference for the time period; Based on the user preference degree of each time period, by comparing the preference degrees of the differentiated time periods, the time period with the greatest preference degree is selected to obtain a target time period selection result.

7. The user music recommendation method based on data analysis according to claim 6, characterized in that: The steps for obtaining the music recommendation result are: Based on the target time period selection result, extract the unplayed music in the popular music list, recommend music to the user, and extract the number of times, frequency and sharing data of the user for the recommended music to obtain music acceptance related data; Based on the music receiving associated data, by the formula: ； Calculate the user's acceptance of the recommended music; in, is the user's acceptance of the recommended music. The number of times the user plays the recommended music. The frequency with which users play the recommended music. The number of times users share the recommended music. , and is the weight parameter; Based on the user's acceptance level of the recommended music, it is compared with a preset acceptance threshold. If the acceptance level is higher than the preset acceptance threshold, the user is marked as being able to receive the recommended music, and music recommendations are made according to the number of times the music of the same music producer is played to obtain a music recommendation result.

8. A user music recommendation system based on data analysis, characterized in that: According to the user music recommendation method of data analysis according to any one of claims 1 to 7, the system comprises: The user preference evaluation module extracts the playing frequency and number of times of each piece of music based on the user's music playing records, evaluates the user's preference for each type of music, analyzes the user's preference for each music tag, and obtains the user preference analysis results; The user classification module extracts the preference data of each user for differentiated music types and music tags based on the user preference analysis results, extracts the user's age and gender data, evaluates the similarity between users, and classifies the users to obtain user classification results; The popular music identification module evaluates the popularity of each piece of music among users of the current category based on the user classification results and the number of times and frequency of each piece of music played by users of the same category, sorts the popularity of the music, and obtains a popular music list; The preferred time period analysis module is based on the user's music playing records and analyzes the number of times the user plays music in different time periods to identify the user's preferred music playing time period and obtain the target time period selection result; The music recommendation module extracts the unplayed music in the popular music list based on the target time period selection results and the popular music list, recommends music to the user, and analyzes the number of times, frequency and sharing data of the recommended music by the user, evaluates the user's acceptance of the recommended music, and obtains the music recommendation results.

Images