Listener screening method and device, storage medium and electronic equipment

Abstract

The invention relates to the technical field of data processing, in particular to an audience screening method, an audience screening device, a storage medium and electronic equipment. The audience screening method comprises the following steps: acquiring listening data of listening to music works of a target singer by a user in a preset period; counting whether the user has a listening behavior in each granularity interval according to a preset granularity based on the listening data, so as to divide the granularity interval into a listening interval and a listening interval; and when the number of the listening intervals exceeds a preset threshold value, marking the user as a listener. According to the audience screening method provided by the invention, the problem that audience evaluation of singers is easily influenced by short-term fluctuation can be avoided.

Description

Technical Field

[0001] The present disclosure relates to the technical field of data processing, and in particular to an audience screening method, an audience screening device, a storage medium and an electronic device. Background Art

[0002] In the current market, several mainstream music platforms such as Spotify, Apple Music, and Tencent Music Entertainment Group (TME) are trying to evaluate the influence and popularity of singers through user behavior data.

[0003] However, existing evaluation methods all have certain flaws. For example, the method used by Spotify is sensitive to short-term market feedback, but due to the short time window, Spotify's model is easily affected by short-term fluctuations and it is difficult to effectively reflect users' long-term interest and loyalty to singers; Apple Music's recommendation system performs well in the short term, but it is overly dependent on algorithms and tends to ignore users' long-term loyalty and behavioral differences; TME's model relies on the number of fans and chart rankings, which makes its evaluation results easily affected by short-term market fluctuations and changes in user behavior, and it is difficult to accurately reflect the long-term value of singers.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0005] The purpose of the present disclosure is to provide an audience screening method, an audience screening device, a storage medium and an electronic device, aiming to solve the problem that the audience evaluation of singers is easily affected by short-term fluctuations.

[0006] Other features and advantages of the present disclosure will become apparent from the following detailed description, or may be learned in part by the practice of the present disclosure.

[0007] According to one aspect of an embodiment of the present disclosure, there is provided an audience screening method, comprising:

[0008] Acquire listening data of music works of a target singer listened by the user within a preset period;

[0009] Based on the listening data, counting whether the user has a listening behavior in each granularity interval according to a preset granularity, so as to divide the granularity interval into a listening interval and a listening interval;

[0010] When the number of the listening intervals exceeds a preset threshold, the user is marked as a listener.

[0011] According to some embodiments of the present disclosure, based on the above solution, the method further includes determining the preset threshold, and the determining the preset threshold includes:

[0012] Acquiring sample listening data of at least one sample user listening to music works of a sample singer during an initial period;

[0013] Traversing each of the sample users, and counting whether the sample users have listening behavior in each granularity interval based on the sample listening data according to a preset granularity, so as to divide the granularity interval into a listening interval and a non-listening interval;

[0014] The preset period is determined based on the non-listening interval of each of the sample users, and a preset threshold is determined based on the preset period and the listening interval.

[0015] According to some embodiments of the present disclosure, based on the aforementioned solution, determining the preset period based on the non-listening interval of each of the sample users includes:

[0016] For one of the sample users, the continuous non-listening intervals are integrated into a continuous non-listening interval, and the number of the non-listening intervals is configured as the interval length of the continuous non-listening interval;

[0017] Traverse all the sample users and count the occurrence ratios of the interval lengths;

[0018] The preset interval length is determined based on the occurrence frequency ratio of the interval length, and the preset period is determined according to the preset interval length and the preset granularity.

[0019] According to some embodiments of the present disclosure, based on the aforementioned solution, determining the preset threshold based on the preset period and the listening interval includes:

[0020] For a sample user, counting the number of listening intervals within the preset period;

[0021] Traversing all the sample users, and counting the cumulative sample users who exceed the times in each interval within the preset period, so as to obtain the proportion of sample users who exceed the times in each interval;

[0022] The preset interval number of times is determined based on the proportion of the sample users, and the preset threshold is determined according to the preset interval number of times and the preset granularity.

[0023] According to some embodiments of the present disclosure, based on the above solution, determining the preset interval number based on the proportion of sample users includes:

[0024] Determine the number of candidate intervals based on the proportion of sample users;

[0025] Calculate the granular user ratio and user balancing parameter corresponding to each candidate interval number respectively; wherein the granular user ratio is the ratio of the cumulative sample users exceeding the candidate interval number within the preset period to all users within the unit preset granularity; the user balancing parameter is the standard deviation of the cumulative sample users exceeding the candidate interval number within the preset balancing period;

[0026] The preset interval number is determined from the candidate interval numbers according to the granular user ratio and the user balancing parameter.

[0027] According to some embodiments of the present disclosure, based on the above solution, the method further includes: determining whether the user has a listening behavior within the granularity interval, wherein the determining whether the user has a listening behavior within the granularity interval includes:

[0028] When the duration for which the user listens to the music work of the target singer exceeds a preset duration, or the user actively plays the music work, it is determined that the user has a listening behavior; otherwise, it is determined that the user does not have a listening behavior.

[0029] According to some embodiments of the present disclosure, based on the aforementioned scheme, the preset granularity is week, the preset period is 7 weeks, and the preset threshold is 2 times.

[0030] According to a second aspect of an embodiment of the present disclosure, there is provided an audience screening device, comprising:

[0031] An acquisition module, used to acquire the listening data of the music works of the target singer listened by the user within a preset period;

[0032] A statistical module, configured to count whether the user has a listening behavior in each granularity interval based on the listening data according to a preset granularity, so as to divide the granularity interval into a listening interval and a listening interval;

[0033] The marking module is used to mark the user as a listener when the number of the listening intervals exceeds a preset threshold.

[0034] According to a third aspect of the embodiments of the present disclosure, there is provided a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the audience screening method in the above-mentioned embodiment is implemented.

[0035] According to a fourth aspect of an embodiment of the present disclosure, an electronic device is provided, characterized in that it includes: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the audience screening method as in the above-mentioned embodiment.

[0036] The exemplary embodiments of the present disclosure may have some or all of the following beneficial effects:

[0037] In the technical solutions provided in some embodiments of the present disclosure, by analyzing the listening data of users listening to the music works of the target singer within a preset period, the number of the user's overall listening intervals within the preset period is used as a standard to measure whether the user is a listener, and then listeners who meet the standards are screened out from the users. On the one hand, the listening data within the preset period is selected, and the preset period can be freely configured, and the cycle can be extended, and then the audience screening is carried out based on long-term data, which can effectively avoid the singer's audience evaluation being easily affected by short-term fluctuations, making the screening results more accurate; on the other hand, the number of listening intervals is used as the screening criterion, which can be intermittent, rather than just looking at the recent continuous listening results, which can make the screening results more objective, adapt to more personalized listening needs of listeners, and the accuracy of audience mining is further improved.

[0038] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The drawings herein are incorporated into the specification and constitute a part of the specification, showing embodiments consistent with the present disclosure, and together with the specification, are used to explain the principles of the present disclosure. Obviously, the drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0040] Figure 1 A flowchart schematically illustrating an audience screening method in an exemplary embodiment of the present disclosure;

[0041] Figure 2 A schematic diagram of a process for determining a preset period and a preset threshold in an exemplary embodiment of the present disclosure is schematically shown;

[0042] Figure 3 A schematic diagram schematically shows the composition of an audience screening device in an exemplary embodiment of the present disclosure;

[0043] Figure 4 A schematic diagram schematically illustrates a computer-readable storage medium in an exemplary embodiment of the present disclosure;

[0044] Figure 5 A schematic diagram schematically shows the structure of a computer system of an electronic device in an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0045] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be more comprehensive and complete and will fully convey the concept of the example embodiments to those skilled in the art.

[0046] In addition, the described features, structures or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced without one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known methods, devices, implementations or operations are not shown or described in detail to avoid blurring the various aspects of the present disclosure.

[0047] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0048] The flowcharts shown in the accompanying drawings are only exemplary and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps can be decomposed, and some operations / steps can be combined or partially combined, so the actual execution order may change according to actual conditions.

[0049] In the current market, several mainstream music platforms such as Spotify, Apple Music, and Tencent Music Entertainment Group (TME) are trying to evaluate the influence and popularity of singers through user behavior data.

[0050] Spotify mainly analyzes the following: (1) User retention: Spotify currently mainly uses a 28-day time window to define the listeners of a singer. The specific method is to calculate the number of users who have played at least one song of the singer for more than 30 seconds in the past 28 days. The advantage of this method is that it can quickly capture the short-term popularity of the singer on the platform and reflect market dynamics through frequent short-term data updates. Another technology of Spotify is to identify core user groups with high-frequency listening behavior for certain singers through the "Super Listeners" program, and use this data for precise recommendations and personalized marketing. (2) Time window: Spotify uses a 28-day time window to collect and analyze listener data. The purpose of this window design is to capture short-term market fluctuations and quickly reflect changes in user interests. (3) User behavior analysis: Spotify builds an audience model by analyzing the user's playback frequency and playback time, focusing on the number of times the user plays within the time window, and obtains the proportion of core listeners through weighted calculation. Spotify also combines recommendation algorithms to further personalize push notifications for "Super Listeners" to strengthen users' stickiness to specific singers. (4) Data weighting: In model construction, Spotify emphasizes the potential loyalty of high-frequency users to the singer by giving them higher weights. This method helps identify users who continue to pay close attention to the singer.

[0051] Based on the above, the following achievements have been made: (1) Sensitive short-term market feedback: Spotify's approach can effectively capture market changes in the short term and help the platform quickly adjust its promotion strategy. For example, after a new song is released, the platform can use this model to quickly identify high-potential songs and promote them, thereby increasing song playback and user engagement. (2) User group segmentation: Through the analysis of "Super Listeners", Spotify can help artists and music producers connect and interact with their core audiences more effectively, thereby enhancing audience engagement and loyalty, and helping artists improve the influence of their music and brands through precise marketing and content customization.

[0052] However, its flaws are: (1) Short time window: Spotify uses a 28-day time window to define listeners, which results in a high correlation between the number of listeners and the amount of playback, and cannot truly reflect the long-term interests and loyalty of users. Short-term fluctuations in behavior can easily mask the actual recognition of the singer by the user. (2) Ignoring individual differences: The existing model fails to fully consider the differences in user listening behavior. Individual differences such as user listening frequency and listening habits are not reflected in the model, resulting in limited accuracy of the evaluation results.

[0053] Apple Music mainly analyzes the following: (1) Personalized understanding: Apple Music mainly relies on its powerful recommendation system to evaluate the influence of singers. By analyzing users' historical listening data, search behavior, playlist collection and other data, Apple Music has built a personalized recommendation model to evaluate the market performance of singers and make accurate user push. In addition, Apple Music uses indicators such as the number of user plays and song collections to evaluate the popularity of singers and the market potential of songs. (2) Personalized recommendation system: Apple Music uses a personalized recommendation algorithm based on collaborative filtering and deep learning to analyze users' historical behavior data and cluster people based on dimensions such as region, culture, economy, and language. Then, it expands songs for similar groups based on u2u. This recommendation system not only improves the user experience, but also helps new songs and new singers enter the user's field of vision faster. (3) Behavioral data integration: Apple Music integrates users' various behavioral data (such as play, search, collection, etc.) to build a comprehensive evaluation model. By weighting different behavioral data, the model can reflect the overall preferences of users and help the platform identify potential popular singers and songs in the market. (4) Long-term analysis of user behavior: Apple Music’s model also takes into account the long-term nature of user behavior. By analyzing changes in user behavior over different time periods, it evaluates the performance of singers at different stages of their life cycle. This analysis helps the platform optimize singer promotion strategies and improve user stickiness.

[0054] Based on the above, the following achievements have been made: (1) Accurate user recommendations: Through its personalized recommendation system, Apple Music can accurately match users’ music preferences, improve user experience, and help new singers gain exposure faster. This method has achieved remarkable results in promoting new songs and driving market innovation. (2) Improving singers’ market performance: Through the integrated analysis of user behavior data, Apple Music’s evaluation model has successfully helped the platform identify and promote a number of potential singers, improving their market performance and popularity.

[0055] However, its flaws are: (1) Algorithm dependence: Apple Music's recommendation system is highly dependent on algorithms. Although it can accurately recommend singers, it is relatively limited in promoting emerging singers and developing the long-tail market, and it is easy to ignore the long-term loyalty of users to specific singers. (2) Unsatisfactory recommendation effect: For emerging singers who have not yet accumulated a large amount of historical data, the recommendation effect may not be ideal, and it is difficult to quickly increase their popularity and user base in the short term.

[0056] Tencent Music Entertainment Group (TME) mainly analyzes the following: (1) Fan conversion: TME's technical solution focuses on the number of fans, playback volume and user interaction behavior, and strives to comprehensively evaluate the market influence of singers through multi-dimensional data analysis. TME focuses on user interaction behavior (such as likes, comments, and shares) to measure the user's emotional investment in the singer. In addition, TME also monitors and improves the exposure and popularity of singers through various charts (such as hot song charts and new song charts). (2) Multi-dimensional data integration: A preliminary exploration of user characteristics of digital music platforms - Taking Xiami Music as an example, TME integrates multi-dimensional user behavior data in model construction, especially fan growth, playback volume and user interaction data. Through the analysis of these data, TME can identify user behavior patterns and make a comprehensive assessment of the popularity of singers. (3) Fan conversion rate analysis: TME also introduced the concept of fan conversion rate, which calculates the proportion of users who convert from ordinary listeners to fans within a certain period of time. This indicator is used to measure the market appeal of singers and the stability of their fan base. (4) Chart ranking mechanism: TME uses multiple charts (such as daily, weekly, and monthly charts) to reflect singers’ market performance in real time. These charts are based on user playback and interaction data, helping the platform to quickly identify popular songs and singers in the market, thereby concentrating resources for promotion.

[0057] Based on the above, the following achievements have been made: (1) Comprehensive evaluation of singer influence: By integrating multi-dimensional data, TME has successfully built a relatively comprehensive singer influence evaluation system. This system not only focuses on users' immediate behavior, but also combines users' interactive data, which can more accurately reflect the market performance of singers. (2) Improving user interaction and platform activity: TME's technical solution has successfully promoted users' interactive behavior and increased the overall activity of the platform. At the same time, through the ranking mechanism, TME is able to optimize resource allocation strategies and focus on promoting high-potential singers and songs.

[0058] However, its flaws are: (1) Fan count data bias: TME measures the popularity of singers by the number of fans and playback volume, but the lack of fan activity and user enthusiasm may lead to data bias and fail to accurately reflect the actual market influence of singers. (2) Short-term fluctuations: Playback volume and chart rankings are greatly affected by short-term traffic fluctuations. The popularity and trend changes of singers can easily lead to large fluctuations, making it difficult to stably evaluate the long-term value of singers.

[0059] Therefore, compared to the evaluation methods of the above-mentioned competing products, the present disclosure proposes a user behavior modeling method with a longer-term perspective, defining the singer's audience as users who effectively play the singer's songs multiple times over a longer period of time. By considering the user's listening frequency, listening continuity, and initiative, the indicators of the present invention can more truly reflect the singer's core audience group, avoiding data bias caused by short-term behavior, and thus providing a more scientific basis for the singer's life cycle classification and resource allocation. This method not only fills the gap in the prior art in long-term window evaluation, but also makes the model more applicable and sensitive by considering individual differences among users.

[0060] The implementation details of the technical solution of the embodiment of the present disclosure are elaborated in detail below.

[0061] Figure 1 The following is a schematic diagram showing a flow chart of an audience screening method in an exemplary embodiment of the present disclosure. Figure 1 As shown, the audience screening method includes steps S101 to S103:

[0062] Step S101, obtaining listening data of a user listening to a target singer's music works within a preset period;

[0063] Step S102, counting whether the user has a listening behavior in each granularity interval based on the listening data according to a preset granularity, so as to divide the granularity interval into a listening interval and a listening interval;

[0064] Step S103: When the number of the listening intervals exceeds a preset threshold, the user is marked as a listener.

[0065] In the technical solutions provided in some embodiments of the present disclosure, by analyzing the listening data of users listening to the music works of the target singer within a preset period, the number of the user's overall listening intervals within the preset period is used as a standard to measure whether the user is a listener, and then listeners who meet the standards are screened out from the users. On the one hand, the listening data within the preset period is selected, and the preset period can be freely configured, and the cycle can be extended, and then the audience screening is carried out based on long-term data, which can effectively avoid the singer's audience evaluation being easily affected by short-term fluctuations, making the screening results more accurate; on the other hand, the number of listening intervals is used as the screening criterion, which can be intermittent, rather than just looking at the recent continuous listening results, which can make the screening results more objective, adapt to more personalized listening needs of listeners, and the accuracy of audience mining is further improved.

[0066] Below, each step of the audience screening method in this example implementation will be described in more detail with reference to the accompanying drawings and embodiments.

[0067] In step S101, the listening data of the music works of the target singer listened by the user within a preset period is obtained.

[0068] Specifically, when determining the listeners of a singer, it is necessary to perform data analysis based on the listening data of the users, so it is necessary to first prepare the initial data, that is, to obtain the listening data of multiple users listening to a target singer's music works within a preset period, and then identify the real user from multiple users. The preset period is a period configured in advance, and how to determine the preset period will be introduced in detail later.

[0069] In step S102, whether the user has a listening behavior in each granularity interval is counted according to a preset granularity based on the listening data, so as to divide the granularity interval into a listening interval and a non-listening interval.

[0070] Specifically, in order to simplify the amount of data processing, the listening data may be processed according to a certain granularity, such as in units of days or weeks, to determine whether the user has a listening behavior within this period.

[0071] In one embodiment of the present disclosure, judging whether the user has listening behavior within the granularity interval includes: when the duration for which the user listens to the music work of the target singer exceeds a preset duration, or the user actively plays the music work, determining that the user has listening behavior; otherwise, determining that the user does not have listening behavior.

[0072] Specifically, to avoid accidents, when a user plays a piece of music, it is judged that the user has listened to the piece of music only when the effective play of a song is the complete play of the song by the user, or the song playing time is greater than or equal to the preset time. The preset time can also be set according to demand, for example, 30s.

[0073] In a preset period, statistics are made according to the above standards to determine whether the user has listening behavior in each granularity interval. If the user has listening behavior, then this granularity interval is a listening interval; if the user does not have listening behavior, then this granularity interval is a non-listening interval.

[0074] In step S103, when the number of the listening intervals exceeds a preset threshold, the user is marked as a listener.

[0075] Specifically, whether the user is a listener is evaluated based on the number of listening intervals of the user during the preset period. This preset threshold is also preset, and the specific method will be described in detail later.

[0076] It should be noted that when counting the number of listening intervals, the data is counted individually in units of each granularity interval, without considering whether there are continuous listening intervals.

[0077] Through the above method, when judging whether a user is a listener of the singer, the judgment is made based on the user's overall listening situation over a period of time, rather than relying solely on recent short-term data. This can effectively avoid the randomness of audience identification, take into account the user's long-term interest and loyalty to the singer, and reduce the impact of short-term fluctuations on the evaluation.

[0078] It takes a certain amount of time for singers to create songs. The frequency of songs created by different singers varies greatly. The listening habits of different users, such as listening continuity, listening frequency, listening type (active / passive), and whether they have listened recently, also vary greatly. Therefore, the number of singers' listeners should be measured in the long term, and users who occasionally listen to the songs because of the singer's short-term popularity but quickly leave can be excluded. Therefore, through user behavior modeling, the singer's listeners can be defined as long-term, multiple, and effective users. The samples are analyzed, and then the appropriate preset period and preset threshold audience indicator construction method is calculated to screen the audience.

[0079] In this disclosure, it is assumed that long-term, multiple, and effective playback proves that the user has a certain degree of acceptance of the singer and has become the singer's "listener". This is illustrated by a real user case:

[0080] Table 1 Different modes of users 1-7 passively listening to singer A

[0081]

[0082] Table 2 Different modes of users 1-7 actively listening to singer A

[0083]

[0084] Table 1 shows the different listening patterns of users 1-7 for singer A. Each column represents the last day of the week, such as '2023-12-25' represents 7 consecutive days from '2023-12-19' to '2023-12-25', and the value represents the number of effective plays of the singer by the user in the corresponding 7 days. By analyzing the existing historical data, the differences in users' listening behavior for singers can be summarized as follows:

[0085] ① Different listening frequencies: For example, users 4 and 5 listen at a low frequency, while users 3 and 7 listen at a high frequency.

[0086] ② Different listening continuity: For example, users 4 and 5 have long listening intervals, while users 3 and 7 have short listening intervals.

[0087] ③ Different listening initiative: Table 2 shows that users 4 and 5 have actively played the music, while users 1, 2, 3, 6, and 7 have not.

[0088] Users 1-7, who are the real listeners of singer A? How to identify the "listeners" of a singer from a large number of users with different behavioral habits? In addition to differences in individual listening behaviors, users' listening to songs is also affected by individual differences in singers, the nature of the songs, and various external factors. Singers continue to create in their careers, and the time and frequency of song releases are uncontrollable. After the release of a new song, "lost" users may be "recalled". From the perspective of songs, some singers' songs are more "listening-worthy" and suitable as background music for scenes such as driving and studying. They themselves have the nature of being replayed at a high frequency. In addition, external factors such as personalized distribution of algorithms and platform value positioning greatly affect the information received by users.

[0089] According to Occam’s razor, this article hopes to define a concise “audience” metric that:

[0090] a. Truthfully describe the number of users who listened to the singer multiple times within a period of time;

[0091] b. It has strong sensitivity and rises rapidly as the singer's popularity increases;

[0092] c. Ability to eliminate randomness: The popularity of a singer’s songs on the website or the increase in exposure on other platforms will lead to accidental playback by a large number of users, but these users will be lost as the singer’s popularity declines, so they need to be eliminated.

[0093] Therefore, when constructing the audience index, this application follows the assumption of "long-term, multiple, and effective" playback, and needs to solve the problem of "how many times does a user listen to a song within a certain period of time to indicate that the user has developed a sense of the singer." The problem can be equivalently transformed into: when a user has not listened to a singer for a certain period of time, the possibility of listening to the singer again will drop to zero, that is, user churn occurs. Then, an audience index that meets the above properties is constructed.

[0094] Figure 2 The following is a schematic diagram of a process for determining a preset period and a preset threshold in an exemplary embodiment of the present disclosure. Figure 2 As shown, the method further includes determining the preset period and the preset threshold, specifically including:

[0095] Step S201, obtaining sample listening data of at least one sample user listening to a sample singer's music work during an initial period;

[0096] Step S202, traversing each of the sample users, and counting whether the sample users have listening behavior in each granularity interval based on the sample listening data according to a preset granularity, so as to divide the granularity interval into a listening interval and a non-listening interval;

[0097] Step S203: determining the preset period based on the non-listening interval of each of the sample users, and determining a preset threshold based on the preset period and the listening interval.

[0098] First, in step S201, sample data needs to be obtained.

[0099] Singer stratified sampling: Since the user stickiness of singers of different styles and stages may be different, the samples need to cover singers who are new, famous, and famous. Singers who entered the site after June 2023 or whose number of players at that time was less than or equal to 100 were restricted, and the singers were divided into 10 groups. According to the total number of people in each group on the site, 10% of each group was randomly sampled, and a sample of 1,658 singers was obtained.

[0100] Weekly user listening to singer data: collected from June 17, 2023 to June 15, 2024, a total of 53 weeks, the number of times each user listened to each singer per week. A total of 76.345 million users were covered, and 421 million listening data were collected.

[0101] Then, in step S202, data processing is performed. According to whether the user continuously listens to a certain singer in units of weeks, the interval of continuous listening or continuous non-listening is identified, and the number of weeks of continuous listening / non-listening for each user is obtained.

[0102] The following is an example of the data record of user 8 listening to singer B in weeks from June 17 to September 9, 2023.

[0103] Table 3 Weekly data of user 8 listening to singer B

[0104]

[0105]

[0106] The week of 2023-06-17 is recorded as the first week w1, and the i-th week is recorded as w i ,i∈Z * .

[0107] Record whether you listened to it during the week

[0108] Group each week, and record the group as g j ,j∈Z * , assume that j is in set J.

[0109] The first week is grouped into the first group and is labeled g1.

[0110] About Zhou i , i>1, when Add to new group j+1 , otherwise merge into the previous group gj .

[0111] Indicates the type of continuous interval group. When the group is a user continuous listening group, otherwise

[0112] Calculate the number of consecutive weeks in each group and record it as

[0113] Table 3 shows that user 8 listened to a certain singer in the week of 2023-06-17, did not listen to it in the three weeks from 2023-06-24 to 2023-07-08, and listened to it again in the week of 2023-07-15. Therefore, the week of 2023-06-17 is recorded as the first group g1, The week from 2023-06-24 to 2023-07-08 is recorded as Group 2 g2. The weekly record for 2023-07-15 is Group 3 g3, And so on.

[0114] To exclude accidental playback, this article counts the types and lengths of continuous interval groups of listening or not listening for users who played any singer for at least 2 weeks within 53 weeks.

[0115] In one embodiment of the present disclosure, in step S203, the preset period is determined based on the non-listening interval of each of the sample users, including: for one of the sample users, integrating the continuous non-listening intervals into a continuous non-listening interval, and configuring the number of the non-listening intervals as the interval length of the continuous non-listening interval; traversing all of the sample users, and counting the proportion of the number of occurrences of each of the interval lengths; determining the preset interval length based on the proportion of the number of occurrences of the interval lengths, and determining the preset period according to the preset interval length and the preset granularity.

[0116] Take the continuous intervals in which all users have no listening behavior, that is, take Make And count the different length intervals The frequency of occurrence is as follows:

[0117] Table 4 Frequency of occurrence of continuous interval length of users without listening behavior for sample singers

[0118]

[0119] Table 4 shows that when users in the sample play a certain singer within a year, the longest interval between two listening sessions is 8 weeks, accounting for only 0.00005%. More than 98% of users have an interval of <= 4 weeks between two listening sessions, which means that the interval is four weeks or less, and it is impossible to determine user churn. When a user has not listened to a singer for more than 8 weeks, there is a high probability that he has churned.

[0120] According to the above results, the present disclosure selects 7 weeks as the interval for counting the number of listening times, that is, the preset granularity is week, and the preset interval length n=7 is the preset period.

[0121] In one embodiment of the present disclosure, in step S203, after obtaining the preset period, determining the preset threshold based on the preset period and the listening interval includes: for one of the sample users, counting the number of intervals of the listening interval within the preset period; traversing all the sample users, and counting the cumulative sample users who exceed the number of each interval within the preset period to obtain the proportion of sample users who exceed the number of each interval; determining the preset interval number based on the proportion of sample users, and determining the preset threshold according to the preset interval number and the preset granularity.

[0122] Among them, determining the preset interval number based on the proportion of sample users includes: determining the candidate interval number based on the proportion of sample users; respectively calculating the granular user ratio and user balance parameter corresponding to each candidate interval number; wherein the granular user ratio is the ratio of the cumulative sample users exceeding the candidate interval number within the preset period to all users within the unit preset granularity; the user balance parameter is the standard deviation of the cumulative sample users exceeding the candidate interval number within the preset balance period; determining the preset interval number from the candidate interval number according to the granular user ratio and the user balance parameter.

[0123] Specifically, continuing to use the above samples, we compare the number of times each singer was played by users for 2 weeks, 3 weeks, ... 7 weeks or more in the week of September 11, 2023 and the previous 6 weeks, and the proportion of users who had played the song at least once in the 7 weeks.

[0124] Table 5 Percentage of people who listened to the program for 1 to n weeks in the 7 weeks before 2023-09-11

[0125]

[0126] Table 5 shows that the average proportion of users who played the song for 1 week within 7 weeks is 79%, the average proportion of users who played the song for 2 weeks or more is 21%, and the average proportion of users who played the song for 3 weeks or more is 10%. According to the above analysis, it is highly likely that the song was played for only 1 week, while the proportion of users who played the song for 4 weeks or more is less than 10%, which is a small amount. Therefore, it is more reasonable to choose to listen to the song for at least 2 or 3 weeks within 7 weeks. That is, the number of candidate intervals is selected as 2 or 3.

[0127] The ratio of the number of people who listened to the program for n weeks to the number of people who played the program that week is shown in Table 6.

[0128] Table 6 Ratio of the number of people who listened to the game for n weeks to the number of people who played the game in the current week

[0129] Number of people who listened for more than 2 weeks / Number of people who listened during the week: average Number of people who listened for more than 3 weeks / number of people who listened during the week: average 99.5％ 41.1％

[0130] Table 6 shows that the number of people who listened to the program for more than 2 weeks within 7 weeks is almost the same as the number of people who listened to the program that week, while the number of people who listened to the program for more than 3 weeks drops to below 50%.

[0131] Finally, we calculated the standard deviation of the number of people who listened to each singer for n weeks in the first 7 weeks within the 53 weeks, and then calculated the mean of the standard deviation to obtain the standard deviation of each statistic within the 53 weeks, see Table 7.

[0132] Table 7 Standard deviation of the number of people who listened to the program for n weeks within one year

[0133] Standard deviation of the number of people who have been here for more than 2 weeks within one year Standard deviation of the number of people who have been here for more than 3 weeks within one year 1,736 723

[0134] Table 7 shows that the standard deviation of the number of people who listened for more than 2 weeks is larger and the changes are more sensitive.

[0135] The sensitivity of 2-week playback data is higher than that of 3-week data, and its magnitude is basically the same as the number of players in that week. It will reach its peak at the same time as the number of players in that week. At the same time, it retains certain properties of a mid-term indicator and is relatively stable. It strictly avoids occasional events of frequent listening within a week.

[0136] Therefore, 2 weeks is selected as the threshold, and users who have listened to the singer for 2 weeks or more in the past 7 weeks are counted as the singer's listeners. That is, the preset granularity is week, and the preset interval number m=2, and the preset threshold is obtained.

[0137] To summarize, this article defines a singer’s “listeners” as users who have effectively played the singer’s songs once or more per week in at least two of the past seven weeks.

[0138] Based on the above method, by analyzing the objective rules of sample data, appropriate audience index parameters are selected according to the statistical results to screen the audience of the singer. When screening the audience, the audience index parameters such as preset granularity, preset period, and preset threshold are directly judged according to the user's listening data, providing a new way to evaluate the singer's audience. At the same time, the audience screening method provided by the present disclosure can effectively avoid reflecting market dynamics through short-term data and effectively reflect the user's long-term interest and loyalty to the singer.

[0139] In one embodiment of the present disclosure, after the listeners of the singer are screened, further data mining can be performed based on the listener data. For example, statistics can be collected on the size of the listener, the listener conversion rate, the proportion of active listeners to the audience, the proportion of fan listeners to the audience, and the average number of plays per listener.

[0140] The singer's life cycle can also be divided according to the audience. According to the size of the audience and the genre of the singer, a reasonable threshold is used to divide the singer's current stage, which makes it easier to lay out the subsequent resource push strategy. The singer's current stage is, for example:

[0141] ① Newcomer (Purely_Newcomer): In the early stage of the creative life cycle, they have not yet formed a fixed style, the quality of their works is average, and the frequency of releasing songs is uncontrollable. The audience size is small, and the proportion of active behavior and long-term retention is low. After the first listening, the probability of replaying is low, and it is difficult to convert to fans.

[0142] ② Notable_Newcomer: A fledgling newcomer who has received some recognition and has a significantly larger audience size than the previous period. The conversion rate has increased, the proportion of active listeners has begun to increase, the conversion rate has improved slightly, the audience has begun to form a preliminary preference for the singer, and the replay rate has increased.

[0143] ③ Rising Star: The singer is new and has made a name for himself. He has received more recognition and is expected to become the next big star. The audience's recognition and preference for the singer have begun to increase significantly, and the rebroadcast rate has increased significantly.

[0144] ④ Blooming (Blooming_Star): In this stage, the singer has become famous and is beginning to be known by more people. At the same time, he has a certain number of loyal fans and listeners. The rate of listeners turning into fans is steadily increasing, the audience's loyalty to the singer is increasing, and the number of loyal fans is increasing.

[0145] ⑤ Well-known (Recognized_Singer): He has 1-3 representative works that are played for a long time by a large number of users. His reputation is gradually rising and he has become a highly-regarded figure in the music industry.

[0146] ⑥Singer (Proper_Celebrity): A top figure on the site, who has at least one song on the main charts, or a top musician in a vertical genre.

[0147] ⑦Super Star: The singer has already become famous, has high off-site exposure, and has appeared in mainstream media.

[0148] Musicians at different stages of their life cycle have different development priorities. For example, novice musicians need to release more original works to enrich their knowledge base, blooming musicians need 1-2 famous works to continuously expand their audience, and top musicians need to seek opportunities to break out of the circle and maintain high popularity and stable audience growth.

[0149] From the perspective of the platform, strengthening the musician ecosystem requires a reasonable resource allocation strategy and an efficient musician promotion channel. The musician life cycle is also used to guide the allocation and operation of algorithms, playlists, and operational resources.

[0150] Currently, the Musician Growth Center has matched monthly growth tasks for musicians based on their current development shortcomings; the musician support strategy within the site has also specified corresponding traffic support strategies based on the musicians' life cycle.

[0151] Figure 3 The following is a schematic diagram showing the composition of an audience screening device in an exemplary embodiment of the present disclosure. Figure 3 As shown, the audience screening device 300 may include an acquisition module 301, a statistics module 302 and a marking module 303. Among them:

[0152] The acquisition module 301 is used to acquire the listening data of the music works of the target singer listened by the user within a preset period;

[0153] A statistical module 302 is used to count whether the user has a listening behavior in each granularity interval based on the listening data according to a preset granularity, so as to divide the granularity interval into a listening interval and a listening interval;

[0154] The marking module 303 is configured to mark the user as a listener when the number of the listening intervals exceeds a preset threshold.

[0155] According to an exemplary embodiment of the present disclosure, the audience screening device 300 also includes a preset module for determining the preset period and the preset threshold, including: obtaining sample listening data of at least one sample user listening to the music work of a sample singer during an initial period; traversing each of the sample users, and counting whether the sample users have listening behavior in each granularity interval according to a preset granularity based on the sample listening data, so as to divide the granularity interval into a listening interval and a non-listening interval; determining the preset period based on the non-listening interval of each of the sample users, and determining the preset threshold based on the preset period and the listening interval.

[0156] According to an exemplary embodiment of the present disclosure, the preset module is also used to integrate the continuous non-listening intervals into a continuous non-listening interval for a sample user, and configure the number of the non-listening intervals as the interval length of the continuous non-listening interval; traverse all the sample users and count the proportion of the occurrence of each interval length; determine the preset interval length based on the proportion of the occurrence of the interval length, and determine the preset period according to the preset interval length and the preset granularity.

[0157] According to an exemplary embodiment of the present disclosure, the preset module is also used to count the number of listening intervals within the preset period for a sample user; traverse all the sample users and count the cumulative sample users who exceed the number of each interval within the preset period to obtain the proportion of sample users who exceed the number of each interval; determine the preset interval number based on the proportion of sample users, and determine the preset threshold according to the preset interval number and the preset granularity.

[0158] According to an exemplary embodiment of the present disclosure, the preset module is also used to determine the number of candidate intervals based on the proportion of sample users; respectively calculate the granular user ratio and user balance parameter corresponding to each candidate interval number; wherein the granular user ratio is the ratio of the cumulative sample users exceeding the candidate interval number within the preset period to all users within the unit preset granularity; the user balance parameter is the standard deviation of the cumulative sample users exceeding the candidate interval number within the preset balance period; the preset interval number is determined from the candidate interval number according to the granular user ratio and the user balance parameter.

[0159] According to an exemplary embodiment of the present disclosure, the statistical module 302 is also used to determine whether the user has listening behavior within the granularity interval, including: when the duration of the user listening to the music work of the target singer exceeds a preset duration, or the user actively plays the music work, it is determined that the user has listening behavior; otherwise, it is determined that the user does not have listening behavior.

[0160] According to an exemplary embodiment of the present disclosure, the preset granularity is a week, the preset period is 7 weeks, and the preset threshold is 2 times.

[0161] The specific details of each module in the above-mentioned audience screening device 300 have been described in detail in the corresponding audience screening method, so they will not be repeated here.

[0162] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0163] In an exemplary embodiment of the present disclosure, a storage medium capable of implementing the above method is also provided. Figure 4 A schematic diagram schematically illustrates a computer-readable storage medium in an exemplary embodiment of the present disclosure, such as Figure 4As shown, a program product 400 for implementing the above method according to an embodiment of the present disclosure is described, which can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can be run on a terminal device, such as a mobile phone. However, the program product of the present disclosure is not limited thereto. In this document, a readable storage medium can be any tangible medium containing or storing a program, which can be used by or in combination with an instruction execution system, an apparatus or a device.

[0164] In an exemplary embodiment of the present disclosure, an electronic device capable of implementing the above method is also provided. Figure 5 A schematic diagram schematically shows the structure of a computer system of an electronic device in an exemplary embodiment of the present disclosure.

[0165] It should be noted that Figure 5 The computer system 500 of the electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0166] like Figure 5 As shown, the computer system 500 includes a central processing unit (CPU) 501, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 502 or the program loaded from the storage part 508 to the random access memory (RAM) 503. In the RAM 503, various programs and data required for system operation are also stored. The CPU 501, the ROM 502 and the RAM 503 are connected to each other through a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0167] The following components are connected to the I / O interface 505: an input section 506 including a keyboard, a mouse, etc.; an output section 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as needed. A removable medium 511, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 510 as needed so that a computer program read therefrom is installed into the storage section 508 as needed.

[0168] In particular, according to an embodiment of the present disclosure, the process described below with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication section 509, and / or installed from a removable medium 511. When the computer program is executed by a central processing unit (CPU) 501, various functions defined in the system of the present disclosure are executed.

[0169] It should be noted that the computer-readable medium shown in the embodiment of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by an instruction execution system, device or device or used in combination with it. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, wherein a computer-readable program code is carried. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, which may send, propagate, or transmit programs for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0170] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0171] The units involved in the embodiments described in the present disclosure may be implemented by software or hardware, and the units described may also be arranged in a processor. The names of these units do not, in some cases, constitute limitations on the units themselves.

[0172] As another aspect, the present disclosure further provides a computer-readable medium, which may be included in the electronic device described in the above embodiment; or may exist independently without being assembled into the electronic device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by an electronic device, the electronic device implements the method described in the above embodiment.

[0173] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0174] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a touch terminal, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0175] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not disclosed in the present disclosure.

[0176] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for selecting an audience, characterized in that: include: Acquire listening data of music works of a target singer listened by the user within a preset period; Based on the listening data, counting whether the user has a listening behavior in each granularity interval according to a preset granularity, so as to divide the granularity interval into a listening interval and a listening interval; When the number of the listening intervals exceeds a preset threshold, the user is marked as a listener.

2. The audience screening method according to claim 1, characterized in that: The method further includes determining the preset period and the preset threshold, wherein determining the preset period and the preset threshold includes: Acquiring sample listening data of at least one sample user listening to music works of a sample singer during an initial period; Traversing each of the sample users, and counting whether the sample users have listening behavior in each granularity interval based on the sample listening data according to a preset granularity, so as to divide the granularity interval into a listening interval and a non-listening interval; The preset period is determined based on the non-listening interval of each of the sample users, and a preset threshold is determined based on the preset period and the listening interval.

3. The audience screening method according to claim 2, characterized in that: The determining the preset period based on the non-listening interval of each of the sample users includes: For one of the sample users, the continuous non-listening intervals are integrated into a continuous non-listening interval, and the number of the non-listening intervals is configured as the interval length of the continuous non-listening interval; Traverse all the sample users and count the occurrence ratios of the interval lengths; The preset interval length is determined based on the occurrence frequency ratio of the interval length, and the preset period is determined according to the preset interval length and the preset granularity.

4. The audience screening method according to claim 2, characterized in that: The determining the preset threshold based on the preset period and the listening interval includes: For a sample user, counting the number of listening intervals within the preset period; Traversing all the sample users, and counting the cumulative sample users who exceed the times in each interval within the preset period, so as to obtain the proportion of sample users who exceed the times in each interval; The preset interval number of times is determined based on the proportion of the sample users, and the preset threshold is determined according to the preset interval number of times and the preset granularity.

5. The audience screening method according to claim 4, characterized in that: The determining of the preset interval times based on the proportion of sample users includes: Determine the number of candidate intervals based on the proportion of sample users; Calculate the granular user ratio and user balancing parameter corresponding to each candidate interval number respectively; wherein the granular user ratio is the ratio of the cumulative sample users exceeding the candidate interval number within the preset period to all users within the unit preset granularity; the user balancing parameter is the standard deviation of the cumulative sample users exceeding the candidate interval number within the preset balancing period; The preset interval number is determined from the candidate interval numbers according to the granular user ratio and the user balancing parameter.

6. The audience screening method according to claim 2, characterized in that: The method further includes: determining whether the user has a listening behavior within the granularity interval, wherein determining whether the user has a listening behavior within the granularity interval includes: When the duration for which the user listens to the music work of the target singer exceeds a preset duration, or the user actively plays the music work, it is determined that the user has a listening behavior; otherwise, it is determined that the user does not have a listening behavior.

7. The audience screening method according to claim 2, characterized in that: The preset granularity is week, the preset period is 7 weeks, and the preset threshold is 2 times.

8. An audience screening device, characterized in that: include: An acquisition module, used to acquire the listening data of the music works of the target singer listened by the user within a preset period; A statistical module, configured to count whether the user has a listening behavior in each granularity interval based on the listening data according to a preset granularity, so as to divide the granularity interval into a listening interval and a listening interval; The marking module is used to mark the user as a listener when the number of the listening intervals exceeds a preset threshold.

9. A computer-readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, the audience screening method according to any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, enables the one or more processors to implement the audience screening method according to any one of claims 1 to 7.

Images