We are now ready to apply the previous graph clustering method to the cluster music songs, according to the tags attached to each song. Alternatively, a dataset of the song playlists can also be used to cluster songs that are often played in many lists. The datasets that we are going to work with can be downloaded from http://www.cs.cornell.edu/~shuochen/lme/data_page.html. The datasets consist of the following files:
train.txt
: This file contains the playlist data by using the integer ID to represent songstags.txt
: This file includes the social tags by using the integer ID to represent songssong_hash.txt
: This file maps a song ID to its title and artisttag_hash.txt
: This one maps a tag ID to its name
Each file has a particular format as explained here:
Format of the playlist data: The first line of the data file consists of the IDs (not the integer ID, but the IDs from other sources for identifying the songs) for the songs, separated by a space...