A subset of the twitter100m_tweets dataset from enryu43:
https://huggingface.co/datasets/enryu43/twitter100m_tweets
This subset contains 1000 tweets from the top 1000 authors from the above dataset. All tweets have also been assigned a sentiment score (1 for positive, 0 for neutral, -1 for negative), using VADER Sentiment Analysis:
https://github.com/cjhutto/vaderSentiment.
The dataset is well suited for both sentiment analysis and authorship identification / obfuscation.
If you use this… See the full description on the dataset page:
https://huggingface.co/datasets/sjmeis/Twitter_1Kx1K.
