Elamur Research⌕
☰
home / papers / 2004 / Exploiting Coreference Annotations for Text-to-Hypertext Conversion

Exploiting Coreference Annotations for Text-to-Hypertext Conversion

Summary

Converting linear documents into hypertext creates two coreference-related needs: linking mentions of the same entity within and across documents, and making each node understandable by itself when anaphors point to text outside the segment. Working in the HyTex project, the authors argue that standard schemes (TEI, MUC, MATE) fall short because they treat coreference as a subtype of anaphora. The two relations differ formally, and mentions in separate documents can corefer without any anaphoric link. Their proposal is a knowledge-based XML annotation scheme that connects discourse entities in the text to concepts in a domain terminology model, the TermNet. TermNet follows WordNet principles and is encoded as XML Topic Maps. Two expressions count as coreferential when they point to the same TermNet entry, and document-internal anaphora is annotated separately. Examples from a German corpus show that the scheme handles both genuine cross-document coreference and homonymous terms. No quantitative evaluation is reported. The contribution is a design that supports automatic hyperlinking and on-demand antecedent resolution.

Key points

  • The scheme uses three main elements. <discourseEntity> (attributes deID, deType) marks mentions. <corefLink> (deIDRef, tmIDRef) ties a mention to a TermNet/Topic Map entry. <cospecLink> (relType, phorIDRef, antecedentIDRef) encodes in-document anaphora with relation types such as substitution or ident.
  • The theoretical motivation follows van Deemter & Kibble (2001). Coreference is an equivalence relation, while anaphora is irreflexive, nonsymmetrical and nontransitive. Bound anaphora, non-referring NPs and intensional contexts show that the two relations diverge.
  • The data is a German corpus covering two domains, "text technology" and "hypertext research". For example, "Annotation" maps to two distinct Topic Map entries (gloss vs. markup), so its uses in the two domains are not linked as coreferent.
  • For cohesive closedness, unresolved anaphors in a node can be handled by four strategies chosen according to relation type: insertion of the antecedent, linking, expanding the visible context, or deletion.
  • HyTex (Hypertextualisierung auf textgrammatischer Grundlage) started in November 2001. It combines a manually annotated document level with a domain knowledge level and stores hypertext views as extra layers that preserve the original texts.

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

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.

@inproceedings{holler2004exploiting,
  title = {{Exploiting Coreference Annotations for Text-to-Hypertext Conversion}},
  author = {Holler, Anke and Maas, Jan Frederik and Storrer, Angelika},
  booktitle = {Proceedings of the Fourth International Conference on Language Resources and Evaluation (LREC'04)},
  year = {2004},
  url = {https://aclanthology.org/L04-1184/}
}

References

15 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Baldwin, B. & A. Bagga (1998). Coreference as the Foundations for Link Analysis over Free Text Databases. In Proceedings of the COLING-ACL'98 Content Visualization and Intermedia Represenation Workshop. CVIR'98. find
  2. Beißwenger, M., A. Storrer & M. Runte (2004). Modellierung eines Terminologienetzes für das automatische Linking auf der Grundlage von WordNet. In Kunze, C. et al. (Eds.): GermaNet: Anwendungen des deutschen Wortnetzes in Theorie und Praxis. In LDV-Forum 19, 1/2. find
  3. Grishman, R. (1995). MUC-6 Coreference Task Definition. URL: http://www.cs.nyu.edu/cs/faculty/grishman/COtask21.book\_1.htm. ↗
  4. Hirschman, L. & N. Chinchor (1997). MUC-7 Coreference Task Definition. URL: http://www-nlpir.nist.gov/related\_projects/muc/proceedings/co\_task .html. ↗ our card
  5. Fellbaum, C. (Ed.) (1998). WordNet: An Electronic Lexical Database. Cambridge: MIT Press. find
  6. Holler-Feldhaus, A. (in press). Koreferenz in Hypertexten: Anforderungen an die Annotation. In Osnabrücker Beiträge zur Sprachtheorie (OBST). find
  7. Kuhlen, R. (1991). Hypertext : Ein nicht-lineares Medium zwischen Buch und Wissensbank. Berlin: Springer. find
  8. Lenz, E. A., B. Birkenhake & J. F. Maas (2004). Von der Erstellung bis zur Nutzung: Wortnetze als XML Topic Maps. In Kunze, C. et al. (Eds.): GermaNet: Anwendungen des deutschen Wortnetzes in Theorie und Praxis. In LDV-Forum 19, 1/2. find
  9. Lenz, E. A. & A. Storrer (2002). Converting a Corpus into a Hypertext: An Approach Using XML Topic Maps and XSLT. In M. Gonzàles Rodríguez & C. Paz Suarez Araujo (Eds.): LREC 2002: Third International Conference on Language Resources and Evaluation, Vol. II (pp. 432–-436). find
  10. Mitkov, R. (2002). Anaphora Resolution. London: Longman, Pearson Education. find
  11. Poesio, M. (2000). Annotating a corpus to develop and evaluate discourse entity realization algorithms: Issues and preliminary results. In M. Gavrilidou et al. (Eds.): Proceedings of the 2 nd International Conference on Language Resources and Evaluation (pp. 211—218). find
  12. Sidner, C. (1979). Towards a computational theory of definite anaphora comprehension in English discourse. Ph.D. thesis. MIT. find
  13. Sperberg-McQueen, C.M. & L. Burnard (Eds.) (2001). TEI Guidelines for Electronic Text Encoding and Interchange. XML-compatible edition. URL: http://www.tei-c.org/P4X/ ↗
  14. XTM: XML Topic Maps (2001). Specification. URL: http://www.topicmaps.org/xtm/1.0/van ↗
  15. Deemter, K. & R. Kibble (2001). On Coreferring: Coreference in MUC and related annotation schemes. In Journal of Computational Linguistics 26, 4 (pp. 629—637). find our card

↑ back to top