Elamur Research⌕
☰
home / papers / 2010 / Building the Basque PropBank

Building the Basque PropBank

Summary

The paper describes how a new annotation layer is added to the Basque Dependency Treebank (BDT): verb semantic roles following the PropBank/VerbNet model. The authors build on three existing resources. The first is the EADB database with syntactic-semantic frames for 100 Basque verbs. The second is an earlier mapping between Basque and English verbs via the Levin classification. The third is the BDT itself, built on the EPEC corpus. They first automatically matched their Levin classes against the updated classes in PropBank/VerbNet, so as to obtain the corresponding English data for each Basque verb. Arguments with an unambiguous case are tagged by the system itself. For ambiguous cases, it proposes several options, and the final choice is made by a linguist. For this purpose the AbarHitz tool was adapted. It displays PropBank/VerbNet information alongside the sentence and provides the annotator with partially filled arg_info relations. At the time of publication, about 12,000 words had been annotated and annotation guidelines had been drawn up. The work shows how the PropBank model can be transferred to a language with a rich case system. It also lays the groundwork for cross-linguistic studies and for automatic role labeling.

Key points

  • BDT is built on the EPEC corpus: roughly 300,000 words of standard written texts. The first stage covers 100 frequent verbs, while light and modal verbs (egin, izan) are not yet considered.
  • Matching the Levin classes yielded the following distribution: 51 % of classes remained unchanged, 6 % gained a new subclass, 2 % changed class, and 41 % of the verbs are absent from PropBank. The first two groups (57 % of the verbs) cover 46 % of EPEC, and these are annotated first.
  • The arg_info tag has the following fields: the VerbNet/PropBank verb (e.g. go_01), the head verb, the dependent element, valency (Arg0–Arg4, ArgM), the VerbNet role, the EADB role, and selectional restrictions ([±animate], [±count], [±hum]).
  • AbarHitz is written in Java. It runs on Windows, Linux and Unix, visualizes the dependency tree and checks syntax during annotation.
  • About 12,000 words have already been annotated. The authors note that Basque and English draw the line between arguments and adjuncts differently, but consider these discrepancies easy to reconcile.

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: Semantic role labeling

Cited by

6 papers of our library cite it (of the 140673 whose bibliographies we read); 1 of them have a page here, 5 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{aldezabal2010building,
  title = {{Building the Basque PropBank}},
  author = {Aldezabal, Izaskun and Aranzabe, María Jesús and Ilarraza, Arantza Díaz de and Estarrona, Ainara},
  booktitle = {Proceedings of the Seventh International Conference on Language Resources and Evaluation (LREC'10)},
  year = {2010},
  doi = {10.63317/4xhm3mj4dez9},
  url = {https://doi.org/10.63317/4xhm3mj4dez9}
}

References

18 references, 3 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Aduriz I., Aranzabe M.J., Arriola J.M, Atutxa A., Díaz de Ilarraza A., Ezeiza N., Gojenola K., Oronoz M., Soroa A., Urizar R. (2006). Methodology and steps towards the construction of EPEC, a corpus of written Basque tagged at morphological and syntactic levels for the automatic processing. In Andre… find
  2. Agirre E., Aldezabal I., Etxeberria J., Pociello E. (2006). A Preliminary Study for Building the Basque PropBank. Proceedings of the 5th International Conference on Language Resources and Evaluations (LREC). Genoa, Italy. find our card
  3. Aldezabal I. (1998). Levin’s verb classes and Basque. A comparative approach. UMIACS Departmental Colloquia. University of Maryland. find
  4. Aldezabal, I. (2004). Aditz-azpikategorizazioaren azterketa. 100 aditzen azterketa zehatza, Levin (1993) oinarri harturik eta metodo automatikoak baliatuz. Leioa (Bilbao): University of Basque Country thesis. find
  5. Aldezabal I., Aranzabe M.J., Arriola J.M., Díaz de Ilarraza A. (2009). Syntactic annotation in the Reference Corpus for the processing of Basque (EPEC): Theoretical and practical issues. Corpus Linguistics and Linguistic Theory 5-2, 245-274. Mouton de Gruyter. find
  6. Baker C.F., Fillmore C.J., Lowe J.B. (1998). The Berkeley FrameNet project. In Proceedings of the COLING-ACL. Montreal, Canada. find our card
  7. Bengoetxea K., Gojenola K. (2007). Desarrollo de un analizador sintáctico-estadístico basado en dependencias para el euskera [Development of a statistical parser for Basque]. Procesamiento del Lenguaje Natural 39, 5-12. find
  8. Bird S., Maeda K., Ma X., Lee H., Randall B., Zayat S. (2002). TreeTrans: Diverse Tools Built on The Annotation Graph Toolkit. Third International Conference on Language Resources and Evaluation, 29-31, Las Palmas, Canary Islands, Spain. find
  9. Civit, M., Aldezabal I., Pociello E., Taulé M., Aparicio J., Màrquez L. (2005). 3LB-LEX: léxico verbal con frames sintácticos-semánticos. In XXI Congreso de la SEPLN. Granada. Spain. find
  10. Figure 1: An example of AbarHitz proposed to the human annotator Díaz de Ilarraza A., Garmendia Aitzpea, Oronoz M. (2004). AbarHitz: An annotation tool for the Basque Dependency Treebank. Paper presented at the International Conference on Language Resources and Evaluation. Lisbon, Portugal. find
  11. Hajic J., Panevová J., Urešová Z., Bémová A., Kolárová V., Pajas, P. (2003). PDT-VALLEX: Creating a Largecoverage Valency Lexicon for Treebank Annotation. In Proceedings of the Second Workshop on Treebanks and Linguistic Theories, 57–68. Sweden. find
  12. Kingsbury P., Palmer M. (2002). From Treebank to PropBank. In Proceedings of the 3rd International Conference on Language Resources and Evaluation (LREC-2002). Las Palmas, Spain. find our card
  13. Levin B. (1993). English Verb Classes and Alternations. A preliminary Investigation. Chicago and London. The University of Chicago Press. find
  14. Marcus M. (1994). The Penn TreeBank: A revised corpus design for extracting predicate argument structure. In Proceedings of the ARPA Human Language Technology Workshop. Princeton, NJ. find
  15. Palmer M., Xue N. (2003). Annotating the Propositions in the Penn Chinese Treebank. In Proceedings of the Second Sighan Workshop, Sapporo, Japan. find
  16. Palmer, M., Gildea, D., Kingsbury, P. (2005). The Proposition Bank: A Corpus Annotated with Semantic Roles. In Computational Linguistics Journal. 31:1. find
  17. Rosén V., Smedt K.D., Dyvik H., Meurer P. (2005). TREPIL: Developing Methods and Tools for Multilevel Treebank Construction. In Civit M., Küber S. and Martí M (eds.), Proceeding of the Fourth Workshop on Trebank and Linguistics Theories, 161-172, Universitat de Barcelona, Spain. find
  18. Xue N. 2008. Labeling Chinese predicates with semantic roles. Computational Linguistics, 34(2): 225-255 find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/aldezabal-2010-building-basque-propbank/