V7.1.0 - enhanced dependencies, consistency overhaul and more

(Note: this version contains the content-identical superset of annotations producing UD_English-GUM in Universal Dependencies V2.8)

• Massive round of consistency corrections and harmonization with English Web Treebank, PTB and OntoNotes
• Added enhanced dependencies
• More error validations
• Added multiword tokens to CoNLL-U format (caution: token IDs like 1-2 now in use!)
• Added reconstructed ellipsis tokens to CoNLL-U format (caution: token IDs like 8.1 now in use!)
• Added metadata to CoNLL-U files
• Better escape characters in Wikification
• ANNIS conversion support for null nodes to accommodate ellipsis tokens

V7.0.0 - new genres, Wikification and more

• 20 documents added from four new genres (total tokens: 150,756):
  • Face to face conversation (material from the Santa Barbara Corpus courtesy of John Du Bois, UCSB)
  • Political speeches (public domain data)
  • Open access text books from OpenStax
  • YouTube Creative Commons-licensed vlogs
• New Wikification layer covering all named entities, including nested and pronominal mentions (work by Yi-Ju Lin)
• Complete overhaul of date/time normalization (work by Nitin Venkateswaran)
• Added function labels to constituent trees
• Added addressee information for speakers in UD data
• Complete overhaul of entity and coreference annotations, incl. separate annotation of split antecedents (work by Yi-Ju Lin and Amir Zeldes)
• Increased consistency with other UD corpora, incl. new and more comprehensive morphological features
• Many corrections to all annotation layers

V6.2.0 - corrections and more consistency

• Massive corrections to entity and coreference annotation
• Massive corrections to TEI date/time annotations by @nitinvwaran
• Removed entity type quantity and manually collapsed to other underlying types
• Added infstat value split for split antecedent (edges are still bridge, but anaphor is now annotated explicitly, not just bridging + giv)
• Coreference infstat and entity type matching now automatically validated (no new with antecedent/given with no antecedent)
• src/tsv/ now contains only coref and bridge edges - all other edge types are derived:
  • Coref chains beginning with a non-first/second person pronoun have an initial cata edge in target/
  • Coref chain links between entities whose heads have the appos dependency in syntax receive the appos coref type automatically
  • Non-cataphoric links emanating from a pronoun receive ana
• Many improvements to syntactic dependency consistency, mostly consolidated with EWT annotation practices

V6.1.0 - corrections and bug fixes

Many corrections and fixes to the build bot

V6.0.0 - first release of GUM series 6

New in this version:

• 22 documents added (total tokens: 129,660)
• Discourse parses in Rhetorical Structure Theory now follow RST-DT guidelines
• 5 new discourse relations (means, manner, attribution, question and same-unit)
• Discourse dependency representation and lisp-style formats available
• Now using native Universal Dependencies syntax trees (not automatic conversion)
• Many manual corrections to lemmatization, POS and other consistency improvements

V5.1.0 - Final release of GUM series 5

Final release of GUM5, numerous corrections

• global overhaul of some lemmas
• fresh constituent parses
• corrections to all annotation layers
• this will be the last version of GUM with Stanford parses as a basis for dependencies (switching to UD as primary gold parses in V6)

V5.0.0 - first release of GUM series 5

New in version 5:

• New documents in academic, bio, fiction and reddit subcorpora
• Split bridging relations into 3 subtypes:
  • bridge:aggr - aggregate reference to multiple antecedents
  • bridge:def - definite entity introduced by bridging
  • bridge:other - all other cases of bridging
• Add morph layer based on UD morphology
• Add sentence type multiple for sentence coordinating multiple types; the type other now only includes sentences not falling into any other category
• Merge Stanford and UD parses for simultaneous queries in ANNIS/PAULA
• Separate coreference and bridging visualizations in ANNIS

V4.2.0 - Final release of GUM series 4

Final release of GUM series 4:

• Added s_type="multiple" for sentences containing multiple types (previously under "other")
• Standardized some @rend from "italics" to always "italic"
• Standardized hyphens/dashes in number ranges to have POS tag 'TO' (e.g. in ranges of years), matching the syntactic analysis
• Changed some inconsistent POS tags for IPA name pronunciation from FW to NP
• Added better imperative mood labeling to CoreNLP UD morph features based on manual s_type annotations
• Removed spurious spans in RST files and fixed some segmentations not conforming to guidelines
• Numerous assorted error corrections

V4.1.0 - corresponds to UD V2.2

Stable release V4.1.0 / V4.1.0nr

• Version number in top level folder suffixed with nr indicates Reddit data is not included in top level folders
• To build the complete V4.1.0 see README_reddit.md (source annotation data is included in _build/src/)
• This version contains the data which was used to generate Universal Dependencies release 2.2 of UniversalDependencies/UD_English-GUM

V4.0.1 - Minor build bot fixes

• Build can be run from other directories with relative path
• utf8 encoding specified for ud conversion
• Annotations identical to V4.0.0