This is the repository for our paper [Identifying Physical Object Use in Sentences](to appear) (EMNLP 2022).
The Toucan Dataset contains 2,123 sentences annotated with one of the Used, Anticipated Use or No Use.
Please fill in the form to request access to the dataset. Note that the sentences were extracted from ICWSM 2009 Spinn3r Blog Dataset. Use of the sentences must abide by the ICWSM Spinn3r Dataset usage agreement. We sparsely extracted sentences from different posts for research analysis purposes only.
For questions and issues, please contact tianyu@cs.utah.edu.