This is a dataset for phrasal verb paraphrasing. This dataset consists of 91 phrasal verbs. Each row gives a phrasal verb, three sentences containing this phrase, and three top paraphrases given by different embeddings.

Annotations: annotators are asked to decide whether these paraphrases are good enough, and they give 1 to a good paraphrase, and give 0 otherwise.

Files: 
1. vp_paraphrasing_annotator_1.xlsx: annotations given by annotator 1;
2. vp_paraphrasing_annotator_2.xlsx: annotations given by annotator 2;
3. vp_paraphrasing_agreed_annotation.xlsx: annotations both annotators agree on.

If you use our data, please cite our work:
Hongyu Gong, Jiaqi Mu, Suma Bhat and Pramod Viswanath, ''Preposition Sense Disambiguation and Representation'', EMNLP 2018.