Weather Sentiment - Amazon Mechanical Turk dataset
Matteo Venanzi, William Teacy, Alexander B. Rogers, Nicholas R. Jennings · ePrints Soton (University of Southampton) · 2015
Dataset re-collected using Amazon Mechanical Turk from an original dataset provided by CrowdFlower as part of the 2013 Crowdsourcing at Scale shared task challenge. The dataset contains 6000 classifications of the sentiment of 300 tweets, with gold-standard sentiment labels, provided by 110 workers. The sentiment judgements are provided in the following categories: negative (0), neutral (1), positive (2), tweet not related to weather (3) and can't tell (4). Each row contains