Raumgeographische Verteilung von Twitter-Hashtags im deutschen Sprachraum

Authors

• Steven Coats University of Oulu

DOI: https://doi.org/10.51814/nm.119533

Keywords:

Twitter, Hashtags, German, spatial analysis

Abstract

This study examines the spatial distribution of hashtags in a corpus of German-language tweets by considering three kinds of user location information: exact location encoded as latitude-longitude coordinates, a „place“ attribute selected from a Twitter-maintained list of places, or a free-form entry in the user profile. Hashtags in tweets with exact locations are slightly more likely to show spatial concentration, compared to hashtags with place or user location information, which may reflect the use of mobile devices to publish tweets. While spatial autocorrelation analysis shows that most hashtags do not exhibit a strong spatial tendency, those that do are likely to be toponyms, appellatives, or proper nouns associated with specific places, as can be shown by mapping autocorrelation values. In addition, some hashtags that exhibit a spatial tendency describe localized geographical or meteorological phenomena.