Elamur Research⌕
☰
home / papers / 2016 / Anaphoricity in Connectives: A Case Study on German

Anaphoricity in Connectives: A Case Study on German

Summary

The paper looks at anaphoric connectives. These are adverbials that find their first argument (Arg1, in PDTB terms) through anaphora resolution rather than through syntactic structure. Many German connectives contain a morpheme that explicitly points backward. Such words can often also act as nominal anaphors, especially relative pronouns, which adds an ambiguity problem.

Using the DiMLex lexicon, the authors counted connectives with anaphoric morphemes and identified those that also have a non-connective reading. They then ran a pilot study on demzufolge. They collected 140 corpus instances, annotated antecedent types, senses and argument spans, and measured inter-annotator agreement. They also tested whether off-the-shelf POS taggers (MATE, ParZu) can separate connective from non-connective uses.

Readings and antecedent types (NP, VP, sentence) turned out to be fairly evenly spread, so a majority baseline is not useful. Annotators agreed only moderately. Taggers solved the disambiguation only partly. The study argues that deeper classifiers are needed, and it maps out a hard subproblem of abstract-anaphora resolution for German.

Key points

  • Of the 274 DiMLex entries, 79 (29%) carry one of 11 anaphoric affixes, such as da-, -dessen and wo-/wes-. Of these 79, 40 are ambiguous between a connective and a nominal-anaphor reading.
  • The demzufolge corpus came from DWDS and has two parts:
    • zeit50: 50 instances from Die Zeit.
    • kernel90: 90 instances from Kernkorpus 20.
  • In kernel90, 33 instances were relative pronouns (non-connective) and 57 were connectives. Antecedents were NP in 47% of cases, VP in 21% and sentential in 32%.
  • On zeit50, three annotators produced fully identical pairs in 69% of comparisons, with sense-level Fleiss-κ = 0.55. Most disagreements concerned whether a use was a connective at all. Senses were mostly cause-result (39 uses), with occasional specialization (4 uses).
  • POS tagging on kernel90 gave mixed results:
    • MATE labelled only 6 of 33 relative pronouns as PRELS.
    • ParZu's KOUS tag identified non-connectives with perfect precision but only 58% recall.
    • ParZu's PROP tag identified connectives with 82% precision and 94% recall.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Topics: Coreference resolution · Part-of-speech tagging

Cited by

4 papers of our library cite it (of the 140673 whose bibliographies we read); 1 of them have a page here, 3 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/w16-0706

@inproceedings{stede2016anaphoricity,
  title = {{Anaphoricity in Connectives: A Case Study on German}},
  author = {Stede, Manfred and Grishina, Yulia},
  booktitle = {Proceedings of the Workshop on Coreference Resolution Beyond OntoNotes (CORBON 2016)},
  year = {2016},
  doi = {10.18653/v1/w16-0706},
  url = {https://doi.org/10.18653/v1/w16-0706}
}

References

16 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Bernd Bohnet. 2010. Top accuracy and fast dependency parsing is not a contradiction. In Proc. of the 23rd International Conference on Computational Linguistics (COLING), pages 89–97, Beijing, China, August. find our card
  2. Stefanie Dipper and Manfred Stede. 2006. Disambiguating potential connectives. In Miriam Butt, editor, Proc. of KONVENS ’06, pages 167–173, Konstanz. find
  3. Stefanie Dipper and Heike Zinsmeister. 2012. Annotating abstract anaphora. Language Resources and Evaluation, 46(1):37–52. ↗
  4. Robert Elwell and Jason Baldridge. 2008. Discourse connective argument identification with connective specific rankers. In Proc. of the IEEE Conference on Semantic Computing (ICSC), Santa Clara/CA. find
  5. Klaus Krippendorf. 2004. Measuring the reliability of qualitative text analysis data. Quality and Quantity, 38(6):787–800. find
  6. Majid Laali and Leila Kosseim. 2014. Inducing discourse connectives from parallel texts. In Proc. of the 25th International Conference on Computational Linguistics (COLING), Dublin/Ireland. find our card
  7. Eleni Miltsakaki, Cassandre Creswell, Katherine Forbes, Aravind Joshi, and Bonnie Webber. 2003. Anaphoric arguments of discourse connectives: Semantic properties of antecedents versus non-antecedents. In Proc. of the 10 Conference of the European Chapter of the ACL, Budapest. find
  8. Renate Pasch, Ursula Brauße, Eva Breindl, and Ulrich Herrmann Waßner. 2003. Handbuch der deutschen Konnektoren. Walter de Gruyter, Berlin/New York. ↗
  9. Emily Pitler and Ani Nenkova. 2009. Using syntax to disambiguate explicit discourse connectives in text. In Proc. of the ACL-IJCNLP 2009 Conference Short Papers, pages 13–16, Suntec, Singapore. find our card
  10. R. Prasad, N. Dinesh, A. Lee, E. Miltsakaki, L. Robaldo, A. Joshi, and B. Webber. 2008. The Penn Discourse Treebank 2.0. In Proc. of the 6th International Conference on Language Resources and Evaluation (LREC), Marrakech, Morocco. find our card
  11. Charlotte Roze, Laurence Danlos, and Philippe Muller. 2012. Lexconn: A French lexicon of discourse connectives. Discours, 10. ↗ our card
  12. Rico Sennrich, Gerold Schneider, Martin Volk, and Martin Warin. 2009. A new hybrid dependency parser for German. In C. Chiarcos, R. Eckart de Castilho, and M. Stede, editors, From Form to Meaning: Processing Texts Automatically. Proc. of the Biennial GSCL Conference 2009, Tübingen. Narr. ↗
  13. Manfred Stede. 2002. DiMLex: A lexical approach to discourse markers. In A. Lenci and V. Di Tomaso, editors, Exploring the Lexicon - Theory and Computation. Edizioni dell’Orso, Alessandria. find
  14. Yannick Versley. 2010. Discovery of ambiguous and unambiguous discourse connectives via annotation projection. In Proceedings of Workshop on Annotation and Exploitation of Parallel Corpora (AEPC), Tartu, Estonia. ↗
  15. Bonnie Webber, Matthew Stone, Aravind Joshi, and Alistair Knott. 2003. Anaphora and discourse structure. Computational Linguistics, 29(4):545–587. ↗ our card
  16. Ben Wellner and James Pustejovsky. 2007. Automatically identifying the arguments of discourse connectives. In Proc. of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning (EMNLP-CoNLL), pages 92–101, Prague, Czech Republic. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/stede-2016-anaphoricity-connectives-case-study-german/