We presented, “Spot the Odd Song Out”: an online game for collecting music similarity data poster at DMRN+7 at Queen Mary University.
Authors:Daniel Wolff(1), Guillaume Bellec(2) and Tillman Weyde(1)
(1) Music Informatics Group, City UniversityLondon.
(2)École Nationale Supérieure de Techniques Avancées, Paris
We present the “Spot the Odd Song Out” online game for collecting music similarity data. Similarity estimation is a key topic in Music Information Retrieval with many applications. In scenarios such as music exploration or recommendation, user satisfaction depends on the agreement between the user and the system on which music is more and which is less similar. The perceived similarity is specific to the individual user and influenced by a number of factors such as cultural background, age, education etc… Our goal is to adapt similarity models to user data, but there are few similarity datasets openly available at this point, and none contains information on user background.
The “Spot the Odd Song Out” game collects relative similarity judgements of users on triplets of songs, where they are asked to choose one song as the “odd song out”. This data is annotated with user attributes such as age, location, language and music taste. The game is designed as multi-player and rewards blind agreement of players.
The game is based on the CASimIR API. Being extensible to multiple question types and scenarios, it manages the storage of user input and user background information, as well as the music presented and the statistical selection of samples. The game has been implemented as a Facebook app. Using social channels for distribution, we hope it will attract a large number of players.
Download the poster here .