GeenStijl.nl embeddings (TGTR-4)
Abstract
Authors
Abstract
We collected over 8M messages from the controversial Dutch websites GeenStijl and Dumpert to train a word embedding model that captures the toxic language representations contained in the dataset. The trained word embeddings (±150MB) are released for free and may be useful for further study on toxic online discourse.
Authors
- Pierre Voué
- Elizabeth Cappon
- Tom De Smedt
Comments are closed.