Abstract
Nowadays music is ubiquitous in contemporary life. People listen to it wherever they are. However, there is the problem that listeners would like to have the right music for their current situation without much effort and interaction. Information systems can support this need by making use of ubiquitous and context-aware music recommendation systems. This paper will focus on the user’s context in form of long-term seasonal factors combined with the music features. Therefore, a music dataset with 3644 songs and 10932 data points of audio features were used and compared with seasonal aspects. The analysis shows seasonal influences on the music consumption. A positive correlation between the music parameters valence as well as energy and the monthly temperature has been detected. In contrast to winter season more active, energetic and happy songs are preferred during summer. The findings can serve to integrate seasonal adjustments to context-aware, ubiquitous music recommender systems.
Recommended Citation
Helmholz, Patrick; Siemon, Dominik; and Robra-Bissantz, Susanne, "Summer hot, Winter not! – Seasonal influences on context-based music recommendations" (2017). ACIS 2017 Proceedings. 6.
https://aisel.aisnet.org/acis2017/6