How digital communication mirrors socio-cultural orientation. A computational sociolinguistic study of regional variation on "Jodel"

We study regional variation in social media communications collected from the chat app "Jodel". Using a corpus of 3.8M discussions from the entire German-speaking area, we combine methods from computational linguistics and sociolinguistics to model and interpret different aspects of socio-pragmatic regionality in online writing. The results show strong correspondences with several socio-cultural spatial structures, such as a dialect division of German, student mobility, work commuting, and region-specific discussion topics. Our study contributes to the emerging field of computational sociolinguistics