Music intelligence universe server
An artificial intelligence song/music recommendation system and method is provided that allows music shoppers to discover new music. The system and method accomplish these tasks by analyzing a database of music in order to identify key similarities between different pieces of music, and then recommends pieces of music to a user depending upon their music preferences.
1. A method of determining a user's preference of music, said method comprising the steps of:
a) providing a digital database comprising a plurality of digital song files;
b) mathematically analyzing each said digital song file to determine a numerical value for a plurality of selected quantifiable characteristics;
c) compiling a song vector comprising a list of said numerical values for each of said plurality of selected characteristic for each said song file;
d) selecting and storing a representative portion of each said song file wherein said representative portion substantially mathematically matches the song vector of said song file;
e) choosing a pair of two dissimilar representative portions and enabling said user to evaluate both representative portions;
f) permitting said user to indicate which of said two dissimilar representative portions said user prefers;
g) based on the indication from said user of which of said two dissimilar representative portions said user prefers, finding another pair of dissimilar representative portions to maximize the probability to learn something about the user's preference; and
h) repeating steps e) through g), as necessary, to establish a taste vector for said user comprising song characteristics that said user prefers.
2. The method according to claim 1 , wherein said method is performed via a computer website.
3. The method according to claim 1 , wherein none of said pairs of two dissimilar representative portions are repeated between consecutive steps.
4. The method according to claim 1 , wherein each of said pairs of two dissimilar representative portions are selected to maximize Euclidian distance between each song represented by said representative portions.
5. The method according to claim 1 , wherein each of said pairs of two dissimilar representative portions are selected to maximize orthogonality with respect to previous pairs of representative portions.
6. A computer implemented method of determining a user's preference of music, comprising the steps of:
providing a digital database comprising a plurality of digital song files:
providing an analysis engine having software for use in a computer processor adapted to execute said software;
using said computer processor to analyze each said digital song file to determine a numerical value for each of a plurality of quantifiable characteristics;
using said computer processor to create a multidimensional song vector for each said digital song file, said multidimensional song vector representing the numerical values for each of said quantifiable characteristics;
selecting and storing a representative portion of each said song file wherein the multidimensional vector for said representative portion substantially mathematically matches the multidimensional song vector of said song file;
said computer processor choosing a pair of two dissimilar representative portions, presenting each said dissimilar representative portions to a user, and enabling the user to evaluate both representative portions;
permitting said user to indicate which of said two dissimilar representative portions said user prefers;
based on the indication from said user of which of said two dissimilar representative portions said user prefers, said computer processor choosing another pair of dissimilar representative portions to maximize the probability to learn something about the user's preference; and
presenting additional pairs of dissimilar representative portions, as necessary, to establish a multidimensional taste vector for said user comprising song characteristics that said user prefers.
7. The method according to claim 6 , further comprising the steps of:
providing a user interface that allows the user to listen to each representative portion.
8. The method according to claim 7 , said user interface further comprising means to enable said user to indicate preference of one representative portion.
9. The method according to claim 6 , wherein said method is performed via a computer website.
10. The method according to claim 6 , wherein none of said pairs of two dissimilar representative portions are repeated between consecutive steps.
11. The method according to claim 6 , wherein each of said pairs of two dissimilar representative portions are selected to maximize Euclidian distance between each song represented by said representative portions.
12. The method according to claim 6 , wherein each of said pairs of two dissimilar representative portions are selected to maximize orthogonality with respect to previous pairs of representative portions.
13. The method according to claim 6 , said analysis engine further comprising a conceptual clustering engine based on physical pattern recognition.
14. The method according to claim 6 , said analysis engine further comprising a non-linear kernel learner.