Datasets:

Size:
n>1T
ArXiv:
License:

Explanation of individual folders

#2
by hac541309 - opened

Reference : https://commoncrawl.org/the-data/get-started/
It seems to generally follow a MM-YY structure, with the caveat that some month slices are combined.
In that case it is MM-MM-YY
I tried to reorder current crawled contents into time ascending order

03-15 CC-MAIN-2015-14 – March 2015
09-16 CC-MAIN-2016-40 – September 2016
11-17 CC-MAIN-2017-47 – November 2017
11-18 CC-MAIN-2018-47 – November 2018
05-19 CC-MAIN-2019-22 – May 2019
05-06-20 CC-MAIN-2020-24 – May/June 2020
10-20 CC-MAIN-2020-45 – October 2020
11-12-21 CC-MAIN-2021-49 – November/December 2021
06-07-22 CC-MAIN-2022-49 – November/December 2022
05-06-23 CC-MAIN-2023-23 – May/June 2023

I propose that the folder names be named with the release names with separate documentation and links to the explanations.

03-15 -> CC-MAIN-2015-14
09-16 -> CC-MAIN-2016-40

this way it would atleast be easier to search for from the common crawl site.

Sign up or log in to comment