Paraphrase Adversaries from Word Scrambling (PAWS)
Paraphrasing IdentificationEnglish
Paraphrase Adversaries from Word Scrambling (PAWS) is a paraphrasing identification-focused dataset in English that provides 750,000+ labeled examples distributed in TSV format.
About Paraphrase Adversaries from Word Scrambling (PAWS)
Dataset contains 108,463 human-labeled and 656k noisily labeled pairs that feature the importance of modeling structure, context, and word order information for the problem of paraphrase identification.
Details
- Task
- Paraphrasing Identification
- Language
- English
- Format
- TSV
- Rows / instances
- 750,000+
- Creator
- Zhang et al.
- Year
- 2019