Hugging Face's logo

Datasets:
lemonilia
/
Elliquiy-Role-Playing-Forums_2023-04

Modalities:
Tabular
Text
Formats:
parquet
Languages:
English
Size:
100K - 1M
Libraries:
Datasets
Dask
License:
Dataset card Viewer Files Community
2
Elliquiy-Role-Playing-Forums_2023-04 / README.md
lemonilia's picture
lemonilia
Update README.md
ca03281
preview code
|
raw
history blame
3.38 kB
metadata
license: other
license_name: agpl-3.0
license_link: https://www.gnu.org/licenses/agpl-3.0.txt
language:
  - en
pretty_name: elliquiy-rp_2023-04
size_categories:
  - 100K<n<1M
configs:
  - config_name: default
    data_files:
      - split: train
        path: elliquiy-rp_2023-04.parquet

Elliquiy roleplaying forum data

A collection of 6,640,593 posts and 112,328 thousands of mostly erotic roleplaying forum threads from Elliquiy, from October 2005 through early April 2023. About 9 GB of uncompressed text data (including formatting tags). The data is from the larger raw Forum RP dataset I also uploaded.

Basic automated cleaning was performed, but the messages are still (by deliberate choice) mostly in HTML format, with the notable exception of converting linebreaks into \n.

In addition to the messages, some metadata was provided for convenience, as well as alternative usernames that could be used instead of directly usernames (in the format of User0, User1 ... UserN). These are unique per-thread, but not globally.

Consider this a work in progress. I might update the dataset in the future as I improve the cleaning procedure.

Limitations

During the scraping procedure (occurred on April 2023) some information like text color and links was lost in the process. Most of the data is adult-themed, and given that usernames inside the posts still exist, it would be best if usernames weren't directly used when training a model with this data.

Given the text formatting used by many users, complete and thorough conversion to Markdown seems very difficult without losing information in the process or causing other issues.

Cleaning procedure details

At the element level

  • Converted post dates to ISO format
  • Simplified blockquotes
  • Removed attributes (class, style) from most tags
  • Changed stylistic div/span into their closest HTML counterpart
    • Some deprecated tags have been used
  • Cosmetic font sizes consolidated into three categories: <small>, normal, <big>
  • Font changes removed
  • Special CSS effects removed
  • Background-colored text into <mark>
  • Spoiler tags converted into <details><summary> blocks
    • However, inline spoilers don't work well with this—to be checked out at a later time
  • Removed left/right "floating text" <div>
  • Removed left/right/justify text-alignment <div>
  • Center alignment elements changed to <center> (deprecated tag)
  • Tried to reduce the amount of <table> elements inappropriately used for cosmetic purposes
    • More work needed

At the text level

  • Removed non-standard unicode spaces
  • Changed generic spoiler text
  • Removed some leftover BB tags (most often the result of user error)
    • More work needed
  • Shortened some bare URLs
  • Tried to turn into <a> links some URLs, when possible (the data was scraped using a forum function that mangled links in some way)
  • Changed elliquiy.com URLs into example.com
  • Removed some site-internal URLs
  • Converted all smilies into emoji
  • Removed excessive newlines and leading/trailing spaces
  • Fixed some HTML element spacing issue
    • More work needed

NOT done

  • Replacing HTML escape characters
  • Changing fancy punctuation to ASCII punctuation
  • Removing usernames entirely
    • Removing them from the message bodies themselves might not be easy.