Analyzing the Potential of Microblogs for Spatio-Temporal Popularity Estimation of Music Artists.

This paper looks into the suitability of microblogs for an important task in music information research, namely popularity estimation of music artists. The research questions addressed are the following: To which extent are microblogs used to communicate music listening behavior? Are there differences between different countries of the world? Is it possible to derive a popularity measure from user’s microblogging activities? We found that microblogging does indeed represent an important communication channel for revealing music listening activities, although the intensity of its use vary considerably from country to country. Motivated by this finding, we took first steps towards a geo-aware, social popularity measure for music artists. To this end, we analyzed user posts mined from the microblogging service Twitter over a period of five months. Addressing the problem of determining the popularity of music artists, we employed a gazetteer on extracted posts relevant for particularmusic artists. The presented approach aims at extracting time- and location-specific artist popularity information. We evaluated the performance of the approach by comparing the popularity rankings derived from Twitter posts against the popularity rankings provided by last.fm, a popular music information system and recommender engine.