A SYSTEM THAT GENERATES INSTANT POPULARITY AND PERSONALIZED RECOMMENDATIONS THROUGH REAL-TIME LISTENING TELEMETRY ON DIGITAL MUSIC PLATFORMS.
Patent Information
- Application Number
- TR202612120
- Authority / Receiving Office
- TR · TR
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2026-07-21
- Publication Date
- 2026-09-21
Smart Images
Figure 00000017_0000
Abstract
Description
- 1 - TARIFF REAL-TIME LISTENING ON DIGITAL MUSIC PLATFORMS INSTANT POPULARITY AND PERSONALIZED RECOMMENDATIONS THROUGH TELEMETRY PRODUCTION SYSTEM 5 TECHNICAL FIELD The invention relates to digital audio streaming systems and real-time data processing. data processing), telemetry analysis, user behavior modeling, machine 10 machine learning, privacy-protected data aggregation, fake traffic (fraud) filtering and personalized digital content delivery, digital music real-time collection and processing of listening events on platforms, Generating real-time popularity scores and providing user decision support. 15 digital music platforms developed for presenting dynamic indicators Real-time listening telemetry provides instant popularity and personalized recommendations. The system that produces the music is related to the invention. The system that is the subject of the invention is digital music services, media streaming. platforms, real-time data streaming infrastructures, and user-specific discovery engines developing software, telecommunications and cloud computing applications It is available. 20 PREVIOUS TECHNIQUE Music recommendations on current digital music platforms are generally based on the past. listening data, total listening numbers, algorithmic suggestions, editorial 25 They are based on lists or weekly popularity rankings. These structures are instantaneous. listening intensity, short-term trend momentum, regional their aggregates, bot / fraud filtering, and privacy-protected aggregation It does not address the issue in an integrated manner; it normalizes it with shifting time windows. unique active listener account, listening event data deduplication mechanisms, 30 Real-time trend score generation, combined with a personalized fit score for decision support. It does not include elements such as layering and thresholded representation. - 2 - Document number US6182128B1, real-time music distribution. It describes the systems. This system provides digital audio based on user requests. This includes file distribution and basic data collection, and real-time listening. telemetry, single listener account with sliding time windows, bot / fraud filtering, elements such as privacy-protected thresholding and dynamic decision support indicators 5 It does not contain. Document number US10891103B1 describes a music-based social network and “currently It includes "listening" visibility and external API integration. It is based on. However, sliding window analysis, normalized single listener account, Fraud-protected trend score, anonymized aggregation, and proactive instability 10 The decontamination layer is missing. Document number US11082742B2 contains shared listening sessions. It offers personalization based on; however, the platform-wide real-time listener counter, Regional / temporal density analysis and integrated bot filtering. not specified. 15 Document number US20250047927A1, listener in live streams It deals with statistics, and the invention involves telemetry collection, data deduplication, and trend analysis. It does not predict the score and the integrity of the personalized decision support architecture. Current technical documentation presents fragmented approaches; real-time eavesdropping event generation, anonymization, valid eavesdropping detection, bot / fraud filtering, 20 Real-time listener count with sliding time windows, trend score generation, personalized a fusion of fit score and thresholded decision support indicators into a single system It does not demonstrate its integration into a scalable architecture. Therefore, Existing solutions enable the rapid capture of instantaneous trend breaks, and new content... increasing discovery, reducing user hesitation, and achieving GDPR compliance 25 It is unable to provide integrated privacy protection. A BRIEF DESCRIPTION OF THE INVENTION Real-time data processing, telemetry analysis, machine learning, and privacy 30 In the protected aggregation area, listening events on digital music platforms Low-latency data collection, verification, anonymization, real-time listening. - 3 - Calculating the number using sliding time windows, generating trend scores, and The invention concerns digital tools for providing personalized decision support indicators. Real-time listening telemetry on music platforms provides instant popularity and A system that generates personalized recommendations has been developed. The developed system includes: collecting listening events, anonymizing them, and valid listening records. Detection, bot / fraud filtering, metadata matching, real-time listener count calculation, trend score generation, regional / temporal ranking, personalized recommendations, and real-world data. By working through time-based visualization steps, it shows the user "how many people are there right now?" "listening", "rise 38% in the last 10 minutes", "currently popular in Istanbul", "yours" Dynamic decision-making, such as "on the rise among users with similar listening history" 10 It provides support indicators and alleviates uncertainty. The developed system includes sliding time windows, rolling windows, and session windows. window, approximate unique count (HyperLogLog), fraud risk scoring, privacy thresholding (k-like anonymity) and personal fit score and overall trend score integrating elements such as fusion, based on the classic total rest count of 15 It is distinct from other systems. DESCRIPTION OF THE FIGURES Figure 1. Real-Time Listening on Digital Music Platforms 20 System architecture that generates real-time popularity and personalized recommendations through telemetry. The corresponding part numbers shown in the figures are given below. 100. User Device and Music Player Interface 25 110. Listening Event Collection Module 120. Anonymization and Privacy Protection Module 130. Real-Time Data Streaming and Queuing Module 140. Valid Listening Detection Module 150. Bot / Fraud and Anomaly Filtering Module 30 160. Metadata and Ownership Matching Module 170. Real-time Listener Count Calculation Module - 4 - 180. Real-Time Trend Score Calculation Module 190. Regional and Time Window Based Sorting Module 200. Personalized Decision Resolution and Recommendation Module 210. Real-Time Counter and Visualization Service Module 5: Caching, Latency Optimization, and Scalability (220) 230. Reporting, Artist Panel and Ownership Analytics Module DETAILED DESCRIPTION OF THE INVENTION Digital voice streaming systems, real-time data processing, telemetry analysis, 10 machine learning, privacy-protected data aggregation, and personalized digital In the field of content delivery, real-time collection of listening events, a system that enables the processing and generation of personalized recommendations based on real-time popularity. The system requires at least one processor, memory, data storage units, and a graphics processing unit. (GPU), network interfaces, cloud / edge computing resources, and distributed data processing 15 to be run on one or more server systems that include the infrastructure The developed invention focuses on real-time listening on digital music platforms. The system generates real-time popularity and personalized recommendations through telemetry; the user a user device and music player interface (100) with which he listens to music and songs Start, pause, resume, skip, add to favorites, and complete 20 a listening event collection that gathers listening events related to interactions such as these module (110) monitors listening events through anonymous or pseudonymous identities an anonymization and privacy protection module (120) that works on listening events a real-time data stream and queuing module (130) that transmits with low latency, A valid listening test determines whether listening events are valid listening or not. 25 Detection module (140) filters bot, fake traffic or anomalous behavior. bot / fraud and anomaly filtering module (150), listening events song ID, ISRC is a system that matches data by artist, album, genre, release region, and ownership. metadata and ownership matching module (160), a specific time window a real-time listener count calculator that calculates the number of real-time listeners. 30 module (170), real-time trend score from short-term listening changes a trend score calculation module (180) that produces songs regionally, temporally and - 5 - a sorting module that sorts contextually (190), with user preferences A personalized recommendation system that generates suggestions by combining real-time popularity signals. Recommendation module (200), the user with the said popularity and recommendation information a real-time counter and visualization that enables display on the interface service (210), caching, latency optimization and scalability module 5 (220) with reporting, artist panel and ownership analytics module (230) It includes. The invention concerns real-time listening on digital music platforms. The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. User device and music player interface (100); mobile application, web player, 10 via desktop application, smart TV or in-car music application It performs music listening operations and produces listening events. This interface works by processing processor and memory resources to initiate playback, Features such as pause, resume, skip, complete, and add to favorites. It records interactions with a timestamp, event ID, session ID, anonymous 15 It associates the user key with the device ID hash. The invention concerns real-time listening on digital music platforms. The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. Listening event collection module (110); song ID, ISRC (International Standard Recording Code (International Standard Recording Code), artist ID, album 20 identity, anonymized user ID, session ID, playback start and end times. time, listening duration, completion rate, repeat listening status, skipping Information, pause / resume information, device type, application version, connection including fields such as type, country / region / city information, language preference, and timestamp. It collects listening events. The module includes event ID, session ID, and anonymous 25. User key, device ID summary, playback position, event generation time, Event arrival time on server, event sequence number, resend information and by associating it with the data deduplication key, the same listening event can be repeated multiple times. This prevents the counting of events that arrive late due to network latency, ensuring accurate accuracy. placing them within a time window and singularly displaying events from different devices. It normalizes the effect on the user account. - 6 - The invention concerns real-time listening on digital music platforms. The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. Anonymization and privacy protection module (120); user identities are pseudonymized or thresholding similar to k-anonymity by operating through anonymous identities, rounding, spacing, and statistical noise addition methods 5 in practice; a specific song, region, city, group of friends or user if the number of active listeners in the segment remains below a predefined threshold value Instead of giving a precise number, they said "few people are listening," "new activity is emerging," "in the region" Qualitative popularity indicators or range values such as "newly emerging" It offers. 10 The invention concerns real-time listening on digital music platforms. The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. Real-time data streaming and queuing module (130); Apache Kafka, Apache Flink, Such as Spark Streaming, Redis Stream, WebSocket, or Server-Sent Events. It provides low-latency event transmission with technologies; network outage, client 15 duplicate events that may occur due to resending or connection interruption It extracts data based on the data deduplication key and event sequence number. The invention concerns real-time listening on digital music platforms. The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. valid listening detection module (140); listening time threshold of a certain number of seconds 20 exceeding, passing a certain percentage of the song, the session being valid, the same by implementing rules such as not allowing unusual recurring messages from the user in a short period of time only playback events that demonstrate genuine user interest are compared to the number of simultaneous listeners. It includes. The invention concerns real-time listening on digital music platforms. 25 The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. bot / fraud and anomaly filtering module (150); listening time, repeat playback frequency, unusual traffic from the same IP or device, at very short intervals initiated listening, account age, device fingerprint, geographical consistency, user Behavioral history, network behavior, proxy / VPN suspicion, artificial traffic to the same segment 30 such as production and deviations from the normal traffic distribution across the platform. Event confidence score and / or anomaly risk score for each listening event using parameters - 7 - assigning weight to events with high confidence scores, excluding events with low scores. holding or evaluating with low weight, isolation forest (Isolation Forest), density-based anomaly detection, time series deviation analysis, or This is supported by audited fraud classification models. The invention concerns real-time listening on digital music platforms. 5 The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. metadata and ownership matching module (160); listening events song ID with ISRC, UPC / EAN (Universal Product Code / European Article Number - Universal Product Code / European Article Number), album ID, artist ID, genre Information, broadcast region and DDEX (Digital Data Exchange) 10 matching based on metadata fields; different recordings of the same musical work versions (remix, radio edit, live recording, acoustic version) as separate track recordings managing as a team or under a collaborative group, sharing royalties and revenue. It supports their processes. The invention concerns real-time listening on digital music platforms. 15 The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. Instantaneous listener count calculation module (170); sliding time windows, Using the rolling window and session window approaches, the last 30 in seconds, last minute, last 5 minutes, last 15 minutes or last hour windows calculating active listening events; by song, artist, album, genre, region or 20 It maintains separate windows based on user segments, anonymizing users. ID, device ID, session ID, and timestamp of the same user multiple times normalizes listening from multiple devices in terms of individual listener count, HyperLogLog, Count-Min Sketch, or probabilistic data structures like Bloom Filter It calculates the approximate number of unique listeners and, depending on the need for precision, exactly 25. It switches between counting and approximate counting modes. The invention concerns real-time listening on digital music platforms. The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. Real-time trend score calculation module (180); recent short time The number of rests in the window, the rate of increase compared to the previous time window, is 30. Listen completion rate, skip rate, add to favorites, share, play Add to list, listen again, regional concentration, similar users - 8 - Normalization by evaluating the rise in the segment and social sharing signals. weighting of registered active listener count + weighting of short-term growth rate + time decay coefficient + listening completion rate weighting + favorite insertion / sharing weight + regional density coefficient + personal fit coefficient - skip rate penalty - bot / fraud risk penalty formula is used to generate the trend score; 5 machine learning-based learning-to-rank, gradient augmentation, logistic regression, multi-arm bandit algorithms, time series forecasting or learning weights with anomaly detection models, past listening events It reduces its effect with the time decay coefficient. The invention concerns real-time listening on digital music platforms. 10 The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. Regional and time window based sorting module (190); country, city, region, language, It generates separate trend scores based on time zone or user segment; in short Time-windowed density increase, geographic clustering, compared to historical average. Geographic masking methods with deviation and minimum user threshold 15 in practice; in the context of morning, noon, evening, night, weekend or special days It reweights the popularity score accordingly. The invention concerns real-time listening on digital music platforms. The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. Personalized uncertainty resolution and recommendation module (200); user preference 20 genres listened to, songs recently listened to, favorite artists, listening time, location, device type, whether it is in motion, connection condition, using premium / free subscription type and in-platform interaction behaviors general instantaneous trend score and personal fit score (genre preference, recent listening) history, real-time listening behavior of similar user segments, skip rate, 25 (completion rate, favorites rate, regional density, and time of day) By combining them, it generates a personalized recommendation score; “currently most listened to automatic selection such as "play the song", "play a currently popular song according to my taste". it offers options and instant popularity in case the user is undecided. By evaluating the score together with the personal fit score, 30 is automatically recommended. He is playing a song. - 9 - The invention concerns real-time listening on digital music platforms. The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. Real-time counter and visualization service (210); instant listener count, trend score, regional intensity score, reliability score and personal fit score by evaluating live counters, trend arrow, heat map, rise label, regional 5 popularity badge, type-based intensity label, or time window-based graphs It produces; visual badges and dynamic labels (“trending in the last 5 minutes”, “this "Number 1 in its genre", "popular among listeners like you" with the user. It transmits it to the interface. The invention concerns real-time listening on digital music platforms. 10 The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. caching, latency optimization and scalability module (220); very often The songs searched were based on popular artists, regional trending lists, and genre. It keeps the rankings in a short-term cache; the cache duration depends on the song's real-time status. its mobility, frequency of questioning, rate of change in listener numbers, and 15 dynamically determines traffic density according to regional or local traffic density; regional or EDGE performs calculations at data processing points at the central layer. integrating, load balancing, horizontal scaling, event queue segmentation and Low latency even under high traffic thanks to the part ID-based shard structure. It provides. 20 The invention concerns real-time listening on digital music platforms. The system, which generates real-time popularity and personalized recommendations through telemetry, is part of a larger system. reporting, artist panel and ownership analytics module (230); instant and periodic popularity trends are handled by the artist, record label or rights holder. It provides analytical output. 25 The invention concerns real-time listening on digital music platforms. The system generates real-time popularity and personalized recommendations through telemetry for the user. When you start a song, the song is collected by the listening event collection module (110). ID, ISRC, artist ID, anonymized user ID, session ID, Playback start and end time, listening duration, completion rate, skip 30 Information such as device type, region information, timestamp, event ID, and data deduplication. An event packet is generated containing fields such as the key and event sequence number. - 10 - Anonymization and privacy protection module (120) allows you to use this package with a pseudonym or anonymously. Converting to identities using thresholding, rounding, and spacing similar to k-anonymity. and applies statistical noise addition methods. Real-time data Streaming and queuing module (130) events Apache Kafka, Apache Flink or similar It transmits data over infrastructures with low latency and in case of network outages, data remains at 5 It provides deduplication. The valid listening detection module (140) listening exceeding the minimum duration threshold, surpassing a certain percentage of the song, session by evaluating parameters such as validity and unusual repetition behavior It only accepts events that reflect genuine user interest. Bot / fraud and Anomaly filtering module (150) listening time, repetition frequency, geographic consistency, 10 suspicion of proxy / VPN violations, device fingerprinting, and platform-wide deviations. calculating event safety scores using criteria such as isolation forest or similar It filters or reduces the weight of suspicious traffic using models. Metadata and Ownership matching module (160) events based on ISRC, UPC / EAN, DDEX Metadata matches artist, album, genre, and release region information, all within the same 15 different versions of the work (remix, live recording, radio edit) as separate or joint works It is managed under the group. Instant listener count calculation module (170) sliding time windows, rolling windows, or session windows using approaches that utilize intervals from the last 30 seconds to 1 hour for a single active session. or calculating the number of users, such as HyperLogLog, Count-Min Sketch, etc. Performing approximate counts with probabilistic data structures and monitoring multiple devices. It normalizes. Real-time trend score calculation module (180) is active. listener count, short-term growth rate, completion rate, skip rate, favorites adding / sharing, regional density and personal fit coefficients over time combining the degradation coefficient with a weighted formula, machine learning 25 Optimizing weights with models (learning-to-rank, gradient boosting, etc.) and applies bot / fraud risk penalties. Regional and time window based. Sorting module (190) by country, city, type, language, time zone and user segment It generates separate trend scores based on geographical clustering and historical averages. It applies deviation and minimum user threshold methods. 30 Personalized uncertainty resolution and recommendation module (200) based on user type preference, recent listening history, current behavior of similar user segments, - 11 - Overall trend score and personal fit with data such as regional density and time of day. By fusing your score, you get a personalized recommendation score and automatic playback. The options include ("play currently popular songs according to my taste," etc.). Results caching, latency optimization and scalability module (220) It is stored via dynamic cache durations, EDGE calculation and shard 5 It is scaled with its structures and provides real-time counter and visualization service. (210) with live counter, trend arrow, heat map, rise label and badges. It is transmitted to the user interface with low latency. The user's new Interactions (listening, skipping, favoriting, completing) as feedback. returning to the system, models are constantly being updated and the reporting module (230) 10 instant popularity trends as analytical output for artists and rights holders. This closed-loop architecture provides the processor, memory, and data storage units. By operating at high scale on GPU and distributed network resources, millions Simultaneous listening is a reliable, privacy-protected, and anti-manipulation process. It is protected and can operate in real-time. 15 25
Claims
- 12 - SYSTEMS 1. Digital voice streaming systems, real-time data processing, telemetry analysis, machine learning, privacy-protected data aggregation, and personalized digital In the field of content delivery, real-time collection of listening events, 5 a system that enables the processing and generation of personalized recommendations based on real-time popularity. The system requires at least one processor, memory, data storage units, and a graphics processing unit. (GPU), network interfaces, cloud / edge computing resources, and distributed data processing. to be run on one or more server systems that include the infrastructure The developed invention focuses on real-time listening on digital music platforms. 10 It is a system that generates real-time popularity and personalized recommendations through telemetry, Its feature is a user device and music player that the user uses to listen to music. Start, pause, resume, skip, add to favorites with the interface (100) a collection of listening events related to interactions such as insertion and completion Listening event collection module (110), listening events anonymously or pseudonymically 15 an anonymization and privacy protection module that operates through named identities (120), a real-time data stream that transmits listening events with low latency. and the queue module (130) checks whether the listening events are valid listening events. a valid listening detection module (140) that identifies bot, fake traffic or A bot / fraud and anomaly filtering module that filters anomaly behaviors 20 (150), listening events song ID, ISRC, artist, album, genre, release region and metadata and ownership matching that matches ownership data. module (160) counts the number of instantaneous listeners within a specific time window. a module for calculating the number of listeners in real time (170), short term A trend score that generates a real-time trend score from listening changes. 25 calculation module (180), sorts songs regionally, temporally and contextually. a sorting module (190) sorting in real time with user preferences A personalized recommendation that generates suggestions by combining popularity signals. module (200) displays the said popularity and recommendation information in the user interface a real-time counter and visualization service that enables display 30 (210), caching, latency optimization and scalability module (220) - 13 - with reporting, artist panel and ownership analytics module (230) It is characteristic.
2. Real-time listening telemetry on digital music platforms according to Claim 1. It is a system that generates instant popularity and personalized recommendations, and its feature is mobile. app, web player, desktop application, smart TV or in-car 5 Listening events via music application are recorded using event ID, session ID, Generated by associating it with an anonymous user key and device ID hash. It is characterized by including the user device and music player interface (100).
3. Real-time listening telemetry on digital music platforms according to Claim 1. It is a system that generates instant popularity and personalized recommendations, and its feature is; song 10 ID, ISRC, artist ID, album ID, anonymized user ID, Session ID, playback start and end time, listening duration, completion. Rate, hop information, device type, region information, timestamp, data deduplication listening events by associating them with the key and event sequence number The collecting listening event is characterized by containing the collecting module (110). 15 4. Real-time listening telemetry on digital music platforms according to Claim 1. It is a system that generates instant popularity and personalized recommendations, and its feature is; k- anonymity-like thresholding, rounding, spacing, and statistical noise Precise listener identification for small user groups by applying insertion methods Anonymization and privacy 20, which offer qualitative popularity indicators instead of numerical ones. It is characterized by containing a protection module (120).
5. Real-time listening telemetry on digital music platforms according to Claim 1. It is a system that generates instant popularity and personalized recommendations, and its features include: Apache Kafka, Apache Flink, Spark Streaming, Redis Stream, WebSocket or low-latency event transmission with technologies similar to Server-Sent Events 25 real-time data streaming that provides and includes data deduplication mechanisms. and is characterized by containing the queue module (130).
6. Real-time listening telemetry on digital music platforms according to Claim 1. It is a system that generates instant popularity and personalized recommendations, and its features include: listening time, song completion rate, session validity, and the same 30 based on at least one of the user's repetitive listening behavior parameters - 14 - by including the valid listening detection module (140) which determines the valid listening It is characteristic.
7. Real-time listening telemetry on digital music platforms according to Claim 1. It is a system that generates instant popularity and personalized recommendations, and its features include: listening time, unusual repetition frequency, multiple device usage, account age, 5 Device fingerprinting, geographical consistency, network behavior, proxy / VPN suspicion. calculating the event confidence score using at least one of its parameters, and bot / fraud and anomaly that excludes or reduces the severity of suspicious events It is characterized by containing a filtering module (150).
8. Real-time listening telemetry on digital music platforms according to Claim 10 It is a system that generates instant popularity and personalized recommendations, and its features include: Listening events are analyzed using ISRC, UPC / EAN, and DDEX-based metadata, including artist and album data. matching different versions of the same work with genre and publication region information. (remix, live recording, radio edit) directing under separate or joint project groups It is characterized by containing a metadata and ownership matching module (160). 15 9. Real-time listening telemetry on digital music platforms according to Claim 1. It is a system that generates instant popularity and personalized recommendations, and its feature is; scrolling using time windows, rolling windows or session windows seconds, last minute, last 5 minutes, last 15 minutes or last hour 20 by including the instant listener count calculation module (170) which includes at least one. It is characteristic.
10. Real-time listening on digital music platforms according to Claim 1. It is a system that generates real-time popularity and personalized recommendations through telemetry, This feature includes probabilistic filters such as HyperLogLog, Count-Min Sketch, or Bloom Filter. Calculating the approximate number of unique listeners using data structures and the same user's 25 Simultaneous listener count that normalizes listening from multiple devices. It is characterized by containing a calculation module (170).
11. Real-time listening on digital music platforms according to Claim 1. It is a system that generates real-time popularity and personalized recommendations through telemetry, Features include: number of active listeners, short-term growth rate, completion rate, skip 30 rate, favorite rate, sharing rate, regional density coefficient and A weighted formula that generates a score using at least one of the personal fit coefficients. - 15 - by including a real-time trend score calculation module (180) It is characteristic.
12. Real-time listening on digital music platforms according to Claim 1. It is a system that generates real-time popularity and personalized recommendations through telemetry, Its features include machine learning-based sorting, gradient incrementing, and 5-bit sequencing. logistic regression, multi-arm bandit algorithm, time series forecasting, or Real-time trend score that includes at least one anomaly detection method. It is characterized by containing the calculation module (180).
13. Real-time listening on digital music platforms according to Claim 1. It is a system that generates real-time popularity and personalized recommendations through telemetry, 10 Features include country, city, type, language, time zone, user segment, and listening type. regional and time window based scores that produce different scores depending on the context. It is characterized by containing the sorting module (190).
14. Real-time listening on digital music platforms according to Claim 1. It is a system that generates real-time popularity and personalized recommendations through telemetry, 15 Features include: user's genre preference, recent listening history, similar users. segment behavior, the user's current popularity in their area, and Using at least one of the time of day data points, calculate the overall trend score and the personal trend score. by including a personalized recommendation module (200) that combines the fit score It is characterized by... 20 15. Real-time listening on digital music platforms according to Claim 1. It is a system that generates real-time popularity and personalized recommendations through telemetry, This feature provides the user with an instant popularity score and personal information when they are undecided. an automatically suggested song by evaluating the compatibility score together. 25 by including personalized uncertainty resolution and suggestion module (200) It is characteristic.
16. Real-time listening on digital music platforms according to Claim 1. It is a system that generates real-time popularity and personalized recommendations through telemetry, Features include: "number of people currently listening", "currently on the rise", "popular in your region". “Your style is on the rise” and “play the currently most listened-to track” 30 Indicators containing at least one of these expressions, such as live counter, trend arrow, temperature. - 16 - Real-time counter and visualization offering a chart or rise label. It is characterized by including service (210).
17. Real-time listening on digital music platforms according to Claim 1. It is a system that generates real-time popularity and personalized recommendations through telemetry, This feature sets a dynamic cache duration for frequently queried songs, 5 EDGE data processing, load balancing, horizontal scaling, and part ID-based Caching that provides scalability through shard architecture, reducing latency. It is characterized by its optimization and scalability module (220).
18. Real-time listening on digital music platforms according to Claim 1. It is a system that generates real-time popularity and personalized recommendations through telemetry, 10 The feature is that it captures instant popularity trends from the artist, record label, or rights holder. Reporting, artist panel, and ownership that provides analytical output to the party. It is characterized by containing the analytics module (230). 20 30