PAR-MEX is a shared task aimed at addressing the challenge of paraphrase detection in the Spanish language, particularly within the context of Mexican gastronomy. It involves the creation of a specialized corpus that participants use to classify and submit their paraphrase detection results.
Language(s)
Spanish (Mexico)
Year
2022
Domain
Gastronomy
Text types
Gastronomy documents
Annotations
Paraphrase or not
Format
csv
Data access
Public
Data link
NLP Topic
Number of units
10298
Type of units
Sentence pairs
Training set size
7382 sentence pairs
Test set size
2819 sentence pairs
Development set size
97 sentence pairs
