AIMP Forum
AIMP for Windows => Предложения / Suggestions => Topic started by: wipeout01 on June 03, 2019, 15:53:39
-
Problem: Users download music constantly, and they do not have time to listen every track. Gigas and gigas of downloaded information in mp3/flac format are stored in the hard disks. That information will never be listened because it is impossible, the users would need days and months to listen every track. As a result, the user will miss beautiful tracks of music never listened.
Solution: Predictive selection algorythm
How it works:
1. When you start your AIMP you should to log in, so AIMP can retrieve important users information to learn about their tastes
2. The users will play their favourite tracks and rate them with several stars.
3. AIMP will convert the rated tracks into metadata creating a musical pattern of the song, analyzing the track and converting the song information into important metadata: Genre, Tempo, Musical Notes, etc.
4. AIMP will learn and learn everytime the user rate a song converting it into metadata (basically it will convert the song in a musical pentagram and store the data as metadata). This will create a PATTERN OF THE SONG.
5. After sometime, AIMP will learn what kind of music style prefer the user, it will create internal statistics showing the kind of refrain, chorus, musical notes, etc., that the user prefer. For example: John Doe prefer tracks of rock music, with refrains of no more than 3 minutes in the song, the refrains must contains the following musical notes: in a 80 % C key, 42 % G key, etc.
6. The information with all the tastes of the users will be stored into the hard disk and uploaded to the AIMP server.
7. CLOUD SERVICES: AIMP will store this information to create an intelligence of "COLECTIVE TASTES".
8. Additional information of the user will be stored as well: male, female, age, etc.
9. This information will be linked to the user account.
The exploration function:
10. After some months, AIMP will know very well the tastes of the user.
11. AIMP should have a "SEARCH FOR COOL MUSIC" button. When you press this button, AIMP will start to search all the mp3/flac files into your hard disk and compare them with the patterns stored.
12. AIMP will select the tracks that fit better with your musical preferences, according to the patterns stored.
13. If you are with your cell phone, do not worry, because AIMP will explore the tracks and download your musical preferences from its server (cloud services) so you can explore cool music in every place.
The business and how to get money:
1. AIMP will receive important information from all its users.
2. AIMP will learn from the tastes of the people by creating patterns of the songs.
3. Those patterns are associated to users: gender, age, country, etc.
SELLING REPORTS ON DEMAND TO THE MUSIC INDUSTRY AND TO INDEPENDENT ARTISTS
4. With that information, AIMP will be able to predict music hits on demand.
5. AIMP will be able to sell reports to the music industry by predicting (according the stored information coming from all the users) the number of probabilities a track has, to become a hit in the market.
6. AIMP could invoice those services to the music industries or even to independent artists (for example: a music band have a demo, but they do not know if the people would like it. With this service, AIMP can help, because according all the collected information from the users, it will predict if the track has possibilities to sucess or not).
-
Thank you. I afraid we have no knowledge and resources to implement and support it. Anyway your ideas looks like similar provides by Last.fm (https://www.last.fm/) service. Does you have tried that?