Leveraging Twitter to detect event names associated with a place

Abstract

The way we refer to a place can often differ from official titles assigned by authorities, especially when referenced in informal settings. This results in a collection of vernacular and alternative names shared and used among the general public but hidden from current gazetteers and place name databases. A common source of these alternative names is when large-scale social events occur in established host sites and the geographical region begins to take on the identity of the event. The limitation of current databases creates conflict in efforts to reconcile the unofficial label with a known location. This paper explores the concept of place-event substitution and methods to detect them via social media with the long-term goal of developing a dynamic place name database. The Twitter streaming API is leveraged in an attempt to monitor host sites and discover patterns of changes in place name terminology among the public.

Keywords::

• spatial data mining
• event detection
• place names
• Twitter