TagATune

This is a data set was created for use on the Music Information Retrieval task of auto-tagging music clips. The data is collected from a two-player online game called Tagatune, deployed on the GWAP.com (Games with a purpose) game portal. In this game, two players are given either the same song or different songs, and are asked to enter descriptions appropriate for their given song. After reviewing each other’s descriptions, the players then guess whether they are given the same song or not.

The data set contains:

• 31383 music clips that are 29 seconds long
• Genres include classical, new age, electronica, rock, pop, world, jazz, blues, metal, punk etc.
• Over 10,000 unique tags from tens of thousands of users
• Only tags that are associated with 50 or more songs are included, and all tags were generated by at least 2 players independently.

Data set available online here: [1]

Original Citation

• Input-agreement: a new mechanism for collecting data using human computation games. Edith Law, Luis von Ah. In CHI '09 Proceedings of the 27th international conference on Human factors in computing systems.
• Online Version of Paper: [2]