Skip to content
Permalink
master
Switch branches/tags
Go to file
 
 
Cannot retrieve contributors at this time

Summary

The German UD is converted from the content head version of the universal dependency treebank v2.0 (legacy).

Introduction

The ACL 2013 paper (https://github.com/ryanmcd/uni-dep-tb/blob/master/ACL2013.pdf, McDonald et al.) describes version 1.0 of the corpus, of which there are 2200 train/800 dev/1000 test sentences in German. According to the paper they consist of Reviews and News genres (the news data being from the TIGER Treebank, Reviews presumably from Google).

The subsequent 2.0 release has more data: 14118 train/799 dev/977 test sentences. Some of the sentences in 1.0 turned out to be duplicated across splits, which was fixed for 2.0. There is no indication in the READMEs of where the new German sentences came from.

Based on the above and the mappings in not-to-release/ud-tiger-mapping.txt, it appears that the genres are:

train: Reviews=s1-s1500, News=s1501-s2200, Web=s2201-s14118 By searching for a selection of sentences in the s2201-s14118 range, i.e. the new ones in version 2.0, it looks like they are from Wikipedia and other websites. dev: Reviews=s1-s500, News=s501-s799 test: Reviews=s1-s301, News=s302-s977

Morphology in release 2.0

Universal POS tags were assigned manually, while LEMMA and XPOSTAG were predicted by TreeTagger (first for release 1.4; see Changelog below). Morphological features were assigned using rules based on the values of the other columns (UPOSTAG, XPOSTAG, LEMMA, FORM, DEPREL). Gender, number and case of nouns and their det/amod children are based on the (manual) syntactic annotation, e.g. nsubj => nominative. They should have high precision but lower recall because we did not add them where the context did not provide enough clues (morphological analyzer / lexicon was not used at this stage).

Morphology in release 2.2

The XPOS and FEATS have been updated using the mate parser with Tueba-D/Z. The mate parser model was trained on the first 80% of Tueba-D/Z as converted by an updated version of the TuebaUdConverter that improved the morphological tag extraction. Tueba-D/Z tokens containing hyphens were split into multiple tokens to align with the UD tokenization.

During tagging, `` '' and -- were temporarily normalized to " and - and some new spellings were normalized back to old spellings to be compatible with the Tueba-D/Z model. Tueba-D/Z is fairly homogenous and somewhat dated, so the tagger does not perform as well on user-generated content as on older newspaper text. A detailed analysis is below.

After incorporating the automatic mate POSTAG and FEATS, the UD sentences were aligned with Tiger using the mapping in #13 and LEMMA (for content words) and POSTAG and FEATS (for all tokens) were updated with gold values from Tiger. The only exception was the feature Voice, whose value is passed through from mate since it is not annotated on the aux in Tiger.

Ordinal numbers split into two tokens in UD (e.g., "3 .") were rejoined as in Tiger and compounds were reanalyzed in order to provide consistent FEATS for all subparts. If any subpart of a compound was tagged as NE by the mate parser, the deprel flat was used, otherwise compound.

Mate model analysis:

The accuracy on Tueba-D/Z test data for XPOSTAG is 98.3% and for FEATS as an unanalyzed string is 91.7%. Analyzing the individual morphological tags, the results are:

       Acc.    Prec. Rec.  F1

Case 0.95580 0.996 0.911 0.952 Definite 0.99653 0.986 0.987 0.986 Foreign 0.99835 0.646 0.714 0.678 Gender 0.96150 0.982 0.929 0.955 Mood 0.99554 0.984 0.959 0.972 NumType 0.99988 0.991 0.998 0.995 Number 0.97852 0.995 0.966 0.980 Person 0.99481 0.986 0.966 0.976 Polarity 0.99998 0.999 0.999 0.999 Poss 0.99993 0.995 0.997 0.996 PronType 0.99734 0.998 0.988 0.993 Reflex 0.99991 0.996 0.992 0.994 Tense 0.99567 0.983 0.962 0.972 VerbForm 0.99492 0.993 0.965 0.978 Voice 0.99934 0.934 0.967 0.950

Case and Gender are not unexpectedly the least accurate of the frequent features. Foreign could potentially be removed, although when inspecting the UD instances the precision seems relatively high.

Changelog

2022-05-15 v2.10

  • Fixed tokenization, lemmatization and tagging of ordinal numerals.
  • Added the Degree feature to adjectives.
  • Fixed values of VerbForm where they did not match XPOS.
  • Fixed the Definite feature: it applies only to DET PronType=Art.

2021-11-15 v2.9

  • Fixed UPOS and DEPREL of attributive usages of 'manche'.
  • Fixed UPOS of prepositions based on XPOS==APPR (especially in multi-word named entities).

2020-11-15 Dan Zeman

  • Fixed UPOS of possessives mein, dein, sein, ihr, unser, euer from PRON (or even PROPN) to DET.

2019-11-15 v2.5

  • Google gave permission to drop the "NC" restriction from the license. This applies to the UD annotations (not the underlying content, of which Google claims no ownership or copyright).

2019-05-15 Dan Zeman

  • Fixed numerous bugs found by the new validator.

2018-11-15 Dan Zeman

  • Fixed punctuation using udapy -s ud.FixPunct

2018-04-15 Adriane Boyd

  • Fixed morphology (retagged with Mate, see above), re-synchronized with original Tiger data where applicable.

2017-11-20 Dan Zeman

  • Fixed: copula is AUX.
  • Fixed: capitalization of multi-word tokens at sentence beginning.

2017-04-13 Dan Zeman

  • Removed duplicate sentences from the training data. They were too long to believe that they were naturally occurring duplicates.

2017-03-01 v2.0

  • Converted to UD v2 guidelines (Dan Zeman)

2016-08-21 Dan Zeman

  • Added sentence ids.

  • Added LEMMA and XPOSTAG predicted by TreeTagger with a German model supplied with the tagger and available in Treex (http://ufal.mff.cuni.cz/treex, commit 50ad1fe0b9907ac382cbcda0a0f102602abc21a0). The UPOSTAGs from the original data (assigned manually) were not modified. Some features were also added if they could be derived from the information already present. Features that need a lexicon and/or disambiguation, such as Gender, Number and Case, have only been added if they can be deduced from the (manually annotated) dependency structure, plus a few heuristics (e.g. form equal to lemma often but not always means singular).

    The work was done mainly using the HamleDT::DE::FixUD block, see https://github.com/ufal/treex/blob/master/lib/Treex/Block/HamleDT/DE/FixUD.pm

2015-11-08 Wolfgang Seeker

  • Removed sentences from test due to overlap with dev (sent-no. 6, 8, 79, 80, 88, 108, 109, 118, 152, 154, 164, 167, 190, 191, 195, 206, 215, 220, 229, 247, 295, 346, 451) Removed sentences from dev due to overlap with train (sent-no. 616)
===================================
Universal Dependency Treebanks v2.0
(legacy information)
===================================

=========================
Licenses and terms-of-use
=========================

For the following languages

  German, Spanish, French, Indonesian, Italian, Japanese, Korean and Brazilian
  Portuguese

we will distinguish between two portions of the data.

1. The underlying text for sentences that were annotated. This data Google
   asserts no ownership over and no copyright over. Some or all of these
   sentences may be copyrighted in some jurisdictions.  Where copyrighted,
   Google collected these sentences under exceptions to copyright or implied
   license rights.  GOOGLE MAKES THEM AVAILABLE TO YOU 'AS IS', WITHOUT ANY
   WARRANTY OF ANY KIND, WHETHER EXPRESS OR IMPLIED.

2. The annotations -- part-of-speech tags and dependency annotations. These are
   made available under a CC BY-SA 4.0. GOOGLE MAKES
   THEM AVAILABLE TO YOU 'AS IS', WITHOUT ANY WARRANTY OF ANY KIND, WHETHER
   EXPRESS OR IMPLIED. See attached LICENSE file for the text of CC BY-SA.

Portions of the German data were sampled from the CoNLL 2006 Tiger Treebank
data. Hans Uszkoreit graciously gave permission to use the underlying
sentences in this data as part of this release.

Any use of the data should reference the above plus:

  Universal Dependency Annotation for Multilingual Parsing
  Ryan McDonald, Joakim Nivre, Yvonne Quirmbach-Brundage, Yoav Goldberg,
  Dipanjan Das, Kuzman Ganchev, Keith Hall, Slav Petrov, Hao Zhang,
  Oscar Tackstrom, Claudia Bedini, Nuria Bertomeu Castello and Jungmee Lee
  Proceedings of ACL 2013

=======
Contact
=======

ryanmcd@google.com
joakim.nivre@lingfil.uu.se
slav@google.com
See https://github.com/ryanmcd/uni-dep-tb for more details

=== Machine-readable metadata (DO NOT REMOVE!) ================================ Data available since: UD v1.0 License: CC BY-SA 4.0 Includes text: yes Genre: news reviews wiki Lemmas: automatic UPOS: converted from manual XPOS: automatic Features: automatic Relations: converted from manual Contributors: Petrov, Slav; Seeker, Wolfgang; McDonald, Ryan; Nivre, Joakim; Zeman, Daniel; Boyd, Adriane Contributing: here Contact: zeman@ufal.mff.cuni.cz

(Original treebank contributors: Quirmbach-Brundage, Yvonne; LaMontagne, Adam; Souček, Milan; Järvinen, Timo; Radici, Alessandra)