Replication Data for: Spatio-temporal machine learning analysis of social media data and refugee movement statistics

In our study, we crawled Twitter data through the Twitter Streaming API and the Twitter REST API by requesting explicitly georeferenced tweets. Then, we merged our dataset with Harvard CGA's dataset (see Harvard CGA Geotweet Archive v2.0 for more details). We subsequently spatially and temporally filtered the dataset to the bounding box [8.0°E, 28.2°N, 43.2°E, 50.0°N] in the World Geodetic System 1984 (WGS 84), and to the time interval between January 2015 and December 2016. The final dataset includes 97,653,736 tweet ids. This dataset is the "Geo-Tweets" dataset from the referenced paper that is used to extract spatio-temporal information about refugee movements. More details can be found in the referenced paper. The full tweet objects can be retrieved via the GET statuses/lookup method of the official Twitter API. Tools such as Twarc or Hydrator can use this endpoint and might be helpful.