Elamur Research⌕
☰
home / papers / 2018 / Anaphora Resolution for Twitter Conversations: An Exploratory Study

Anaphora Resolution for Twitter Conversations: An Exploratory Study

Summary

The paper looks at resolving third-person singular pronouns in Twitter conversations, a setting that had not been studied before. Here a conversation is a chain of replies. The authors argue that the genre combines difficulties from three areas: noisy social media text, multi-party dialogue structure, and features borrowed from speech. They gathered English tweets in December 2017 with twarc and rebuilt the reply trees. From each tree they kept a single longest root-to-leaf path. They then selected threads of three or more tweets that contain he/she/it-type pronouns. Tokenization chained Twokenizer and PTBTokenizer. Full coreference chains were annotated in MMAX2, using adapted guidelines from Grishina and Stede (2015). Each pronoun was also labelled as anaphoric, cataphoric or exophoric, and exophoric cases were subdivided, for example by whether the referent sits in an attached image, a quoted tweet or a link. Finally, they ran the off-the-shelf Stanford resolver on the data and analysed its errors qualitatively. The resulting corpus and the catalogue of genre-specific phenomena give a starting point for adapting resolvers trained on standard text.

Key points

  • Corpus composition: 185 threads, 161 selected for he/she forms and 24 for it forms. Threads run from 3 to 78 tweets, with a mean of 10 and a median of 7. In 77 threads the root tweet carries visual media, and 20 contain a quoted tweet.
  • Annotation statistics: 278 coreference chains, 1438 mentions, 853 pronouns and 65 predicative NPs. Chains have a mean length of 5, a median of 3 and a maximum of 56. 100 chains stay within one tweet and 178 span several tweets.
  • Tokenization: Twokenizer output is passed on to PTBTokenizer. The authors prefer over-splitting, because MMAX2 cannot mark a markable smaller than one token.
  • Mention-matching errors of the Stanford system:
    • 39% come from hashtag and @-variants of a name not being linked to the plain name.
    • 23% come from case sensitivity, e.g. all-caps names.
    • 7% come from missing world knowledge.
  • Other Stanford error types:
    • Missing chains make up 12% of all errors.
    • Among span errors, 36% are caused by usernames being included in the mention span.

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

Cited by

10 papers of our library cite it (of the 140673 whose bibliographies we read); 1 of them have a page here, 9 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/w18-0701

@inproceedings{aktas2018anaphora,
  title = {{Anaphora Resolution for Twitter Conversations: An Exploratory Study}},
  author = {Aktaş, Berfin and Scheffler, Tatjana and Stede, Manfred},
  booktitle = {Proceedings of the First Workshop on Computational Models of Reference, Anaphora and Coreference},
  year = {2018},
  doi = {10.18653/v1/w18-0701},
  url = {https://doi.org/10.18653/v1/w18-0701}
}

References

19 references, 5 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Kevin Clark and Christopher D. Manning. 2015. Entity-centric coreference resolution with model stacking. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers), pages… find our card
  2. Adèle Désoyer, Frédéric Landragin, Isabelle Tellier, Anas Lefeuvre, and Jean-Yves Antoine. 2016. Coreference resolution for french oral data: Machine learning experiments with ancor. In 7th International Conference on Intelligent Text Processing and Computational Linguistics (CICLing), Konya, Tu… find
  3. Kevin Gimpel, Nathan Schneider, Brendan O’Connor, Dipanjan Das, Daniel Mills, Jacob Eisenstein, Michael Heilman, Dani Yogatama, Jeffrey Flanigan, and Noah A. Smith. 2011. Part-of-speech tagging for twitter: Annotation, features, and experiments. In Proceedings of the 49th Annual Meeting of the We ob… find
  4. Scoring coreference partitions of predicted mentions: A reference implementation. In Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 30–35, Baltimore, Maryland. Association for Computational Linguistics. Association for Computat… find our card
  5. Yulia Grishina and Manfred Stede. 2015. Knowledge-lean projection of coreference chains across languages. In In Proceedings of the 8th Workshop on Building and Using Comparable Corpora, Beijing, China. Association for Computational Linguistics. find our card
  6. Alan Ritter, Sam Clark, Mausam, and Oren Etzioni. 2011. Named entity recognition in tweets: An experimental study. In Proceedings of the 2011 Conference on Empirical Methods in Natural Language Processing, pages 1524–1534, Edinburgh, Scotland, UK. Association for Computational Linguistics. find
  7. Iris Hendrickx and Vronique Hoste. 2009. Coreference resolution on blogs and commented news. In Anaphora Processing and Applications, Lecture Notes in Artificial Intelligence 5847, pages 43–53. Springer, Berlin/Heidlberg. find
  8. Tatjana Scheffler. 2017. Conversations on twitter. In Darja Fier and Michael Beiwenger, editors, Researching computer-mediated communication: Corpus-based approaches to language in the digital world, pages 124–144. University Press, Ljubljana. find
  9. Prateek Jain, Manav Ratan Mital, Sumit Kumar, Amitabha Mukerjee, and Achla M. Raina. 2004. Anaphora resolution in multi-person dialogues. In Proceedings of the 5th SIGdial Workshop on Discourse and Dialogue, pages 47–50, Cambridge, Massachusetts, USA. Association for Computational Linguistics. find
  10. David Schlangen, Timo Baumann, and Michaela At-terer. 2009. Incremental reference resolution: The task, metrics for evaluation, and a Bayesian filtering model that is sensitive to disfluencies. In Proceedings of the SIGDIAL 2009 Conference, pages 30–37, London, UK. Association for Computational Ling… find
  11. Xiaoqiang Luo, Radu Florian, and Todd Ward. 2009. Improving coreference resolution by using conversational metadata. In Proceedings of Human Language Technologies: The 2009 Annual Conference of the North American Chapter of the Association for Computational Linguistics, Companion Volume: Short Paper… find
  12. Amanda Stent and Srinivas Bangalore. 2010. Interaction between dialog structure and coreference resolution. In Proceedings of the IEEE Spoken Language Technology Workshop, pages 342–347. IEEE. find
  13. Michael Strube and Christoph Müller. 2003. A machine learning approach to pronoun resolution in spoken dialogue. In Proceedings of the 41st Annual Meeting of the Association for Computational Linguistics, pages 168–175, Sapporo, Japan. Association for Computational Linguistics. find our card
  14. Christopher D. Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven J. Bethard, and David Mc-Closky. 2014. The Stanford CoreNLP natural language processing toolkit. In Association for Computational Linguistics (ACL) System Demonstrations, pages 55–60. find our card
  15. Joel Tetreault and James Allen. 2004. Semantics, dialogue, and pronoun resolution. In Proceedings of the SemDial Conference (CATALOG), Barcelona, Spain. find
  16. Christoph Müller and Michael Strube. 2006. Multi-level annotation of linguistic data with MMAX2. In Sabine Braun, Kurt Kohn, and Joybrato Mukherjee, editors, Corpus Technology and Language Pedagogy: New Resources, New Tools, New Methods, pages 197–214. Peter Lang, Frankfurt a.M., Germany. find
  17. Nobal B. Niraula and Vasile Rus. 2014. A machine learning approach to pronominal anaphora resolution in dialogue based intelligent tutoring systems. In Computational Linguistics and Intelligent Text Processing, pages 307–318, Berlin, Heidelberg. Springer. find
  18. M. Poesio, A. Patel, and B. Di Eugenio. 2006. Discourse structure and anaphora in tutorial dialogues: An empirical analysis of two theories of the global focus. Research on Language and Computation, 4(2-3):229–257. find
  19. Sameer Pradhan, Xiaoqiang Luo, Marta Recasens, Eduard Hovy, Vincent Ng, and Michael Strube. 2014. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/akta-2018-anaphora-resolution-twitter-conversations-exploratory/