v1.5

Now includes sequential and tokenised annotations,

Not many changes to actual annotations, but major update to how the annotations are presented etc.:

• Removed a few very long Named Entity annotations (from wind-en and from htfl-en; counts updated) over which there was doubt whether it was a real NE.
• Updated normalisation:
  • Replaced "İ" with "I" in the annotations to avoid problems lowercasing (concerns mainly wind_en_01)
  • Removed rare but problematic characters: ["", "", "", "", "�"] (not handled well by some transformers)
• Major update of README.md
• Different structure of all data:
  • include sequential annotations
  • include tokenised versions of annotations

v1.4 normalised

Identical to version 1.3, except with minor normalisation for both the text files and annotations:
unicodedata.normalize("NFC", text)
normalising all dashes to "-", all single quotes to "'" and all double quotes to '"'

v1.3 Github release

First release on Github, after completion of TermEval shared task.