The WordSim dataset contains word pairs with human annotated similarity scores.
The columns are Word 1, Word 2, Human (Mean).
353 rows of word similarieties.
Although small, this dataset is very popular and have been used in many researches, including fastText (that called this WS353).
This dataset licensed under a Creative Commons Attribution 4.0 International License.
The English WordSim-353 word pairs and instructions are credited to:ā¦ See the full description on the dataset page:
https://huggingface.co/datasets/almogtavor/WordSim353.
