Datasets:
Tasks:
Summarization
Languages:
English
Multilinguality:
monolingual
Size Categories:
1M<n<10M
Language Creators:
crowdsourced
Annotations Creators:
no-annotation
Source Datasets:
original
License:
body vs content?
#4
by
arturomp
- opened
Perhaps silly question - what's the difference between 'body' and 'content' for this dataset? Is it something like
body = content + summary?
Is it always in that sequence? can it be
body = summary + content?
This means that if the TL;DR expression appears in the middle of the text, the part preceding it is considered to be the content and the part after as the summary.
And to answer the previous question (sorry for the late response), body text encompasses content and summary but the order doesn’t matter.
The cleaned content-summary pairs of the dataset are formed after splitting the body text by the various TL;DR patterns.