Through music videos watched on YouTube, downloaded music files and comments about songs on Facebook and Twitter, social media provides a wealth of information that can be valuable to the music industry for judging the popularity of artists and assessing the music taste of the public.
However, the data are mostly unstructured, dispersed and difficult or only with great effort accessible. A project of St. Pölten University of Applied Sciences is developing an analysis programme that should automatically evaluate data from social media and make it useful for the music industry.
Automatic text analysis
Comments in social media are scattered over many sources, submitted by many different people and with regard to content - if it is not simply a "Like" - linguistically only possible to evaluate individually. An analysis involves large amounts of data and is time-consuming.
In the project SAMBA (Smart Data for Music Business Administration), scientists from the Department of Media and Economics, as well as the Department of Media and Digital Technologies at St. Pölten UAS are working on a process that should make the evaluation of comments easier for the music industry.
The researchers use and develop context-based analysis methods and procedures for structuring and visualizing the data. "We also use semantic techniques that can automatically recognize the content of texts, for example, in the framework of a so-called sentiment-analysis, which can distinguish positive opinions and comments from negative ones", explained Andreas Gebesmair, Head of the project as well as the Institute of Media Economics at St. Pölten UAS.
"Existing programs look above all at statistics on views, likes and engagement. Our project goes further and also concentrates on the content of the comments", said Gebesmair.
Project SAMBA (Smart Data for Music Business Administration)
The SAMBA project is financed by the Bundesministerium für Wissenschaft, Forschung und Wirtschaft (Federal Ministry of Science, Research and Economics) within the framework of the COIN-programme line "Aubau". Advisory partners for the project are: representatives from radio 88.6, Universal Music, TU Vienna (Vienna University of Technology), VTMÖ-Verband unabhängiger Tonträgerunternehmen, Musikverlage und MusikproduzentInnen Österreich (Association of Independent Sound Recording Companies, Music Publishers and Music Producers of Austria) as well as JKU Linz (Johannes Kepler University Linz).