V4.0.0 - Initial release of GUM series 4

NOTE: the version in top level folders is numbered 4.0.0nr for "no reddit". The full version 4.0.0 can be generated automatically, see README_reddit.md on compiling reddit data.

• Added 25 new documents in four new genres
  • New genres: academic, bio, fiction, reddit
  • Total: 101 documents
  • The 6 reddit documents are only included without token data in _/build/src/
• Build bot now generates Universal Dependencies parses and morphology from gold Stanford Dependencies
• Numerous error corrections and build bot bug fixes

V3.2.0 - final release of GUM series 3

Added Universal Dependencies version, metadata, numerous corrections

Revised release of GUM 3

• Corrections across all formats
• Validation scripts now provide warnings against many semantically implausible annotations

V3.0.0 - Initial release of GUM series 3

Release of GUM 3.0.0

• Added 22 documents (total now 76 documents, 64,006 tokens)
• Minor improvements to build bot
• Assorted error corrections to existing documents

V2.3.2 - final release of GUM series 2

• Numerous corrections
• Stabilized build bot
• This should be the final release version of GUM 2
• Stay tuned for GUM 3 with more new data!

V2.3.1 - build bot rc2

Release candidate 2 for new auto build process

V2.3.0 - build bot release candidate 1

First release candidate for auto build process

V2.2.0

• Minor corrections
  • Tags for brackets made more consistent: use literal brackets in TT tags, LRB style in PTB tags (see #4 for details)
  • Minor token consistency corrections (some formats were using incorrect ASCII equivalents for Unicode quotation marks and other punctuation)
  • Adjusted sentence border in syntax files for GUM_interview_gaming
  • Missing sentence type added in GUM_news_imprisoned
  • Coref conversion errors fixed in GUM_whow_chicken
  • Numerous minor dependency corrections

V2.1.1

NB: Contains a few known errors involving tagging of brackets (see #4). This version is the basis for:

Zeldes, Amir and Simonson, Dan (2016) "Different Flavors of GUM: Evaluating Genre and Sentence Type Effects on Multilayer Corpus Annotation Quality". In: Proceedings of LAW X – The 10th Linguistic Annotation Workshop. Berlin, 68–78.