Multi-Domain Text Classification Datasets

AbstractOff-the-shelf models are widely used by computational social science researchers to measure properties of text, such as sentiment. However, without access to source data it is difficult to account for …...