. blog —

artist similarity visualization

worldofmusic.jpga data visualization showing 9,276 artists & how they are related to each other. the artist relation data is mined from 150,000 users with 2.5 million ratings of artists in the Yahoo! Music service. this dataset is complex because the relationships between artists are typically divided by a power law distribution, i.e. there are few nodes with a high degree & many nodes with a low degree. a technique is used called 'semidefinite programming' to layout & cluster the data, which maps high dimensional data into a lower dimensional euclidean vector space. see also music plasma for a similar but more interactive approach. [stanford.edu(pdf)|via sun.com]

— Share this Article —