Elamur Research⌕
☰
home / papers / 2018 / Annotating Attribution Relations in Arabic

Annotating Attribution Relations in Arabic

Summary

The authors tackle the absence of any gold-standard resource for attribution, meaning who said or believed what, in Modern Standard Arabic. Earlier Arabic work on attribution was mostly authorship identification on long texts or tweets. It ignored the internal structure of reported speech and opinion in short texts. Building on English and Italian schemes from the PDTB, PARC3 and ItAC, the team adapted annotation guidelines to Arabic news from the Arabic Treebank (ATB). The guidelines cover three core elements: the cue, the source and the content. Each element carries syntactic and semantic features. The scheme handles both direct quotations and indirect speech. Indirect speech is harder because punctuation, speech verbs or complementizers may be missing, which blurs where the source and content begin and end. The authors also built a dedicated annotation tool, ESNAD, so that every attribution instance is captured consistently. They ran a pilot on 20 news articles and analysed agreement between annotators. This is a first step toward the first Arabic attribution gold corpus, a resource useful for information extraction, QA, summarization and opinion mining.

Key points

  • Data: news texts from the Penn Arabic Treebank (ATB, LDC2010T13). The pilot covers 20 articles from several sources.
  • Annotated elements are cue, source and content, each with syntactic and semantic features. Direct and indirect attribution are both in scope.
  • ESNAD is a new attribution annotation tool built specifically for Arabic.
  • The scheme follows the PDTB and the PARC3/ItAC line of work, which annotates attribution regardless of whether a discourse relation is present. For reference, PARC 3.0 has 19,712 attribution relations, and its span agreement is 100% for source, 91% for cue, 94% for content and 46% for supplement.
  • The paper notes that the PDTB-style discourse banks for Chinese, Arabic, Hindi, Czech and Turkish all left attribution out of their schemes.

Interesting ideas

  • I1. Model attribution as its own relation with three slots (cue, source, content) rather than as a by-product of discourse relations, so nested and standalone reports are not lost.
  • I2. Separate direct quotation, where the author does not alter the content, from indirect speech, where markers may be missing and the content is the writer's rendering.
  • I3. The examples show material wedged between source, cue and content, such as relative clauses about the speaker and circumstantial phrases. Annotating these as distinct pieces keeps the content span clean.
  • I4. Measure inter-annotator agreement separately for each element instead of as one score. PARC 3.0 shows source spans are easy while supplement spans are hard.
  • I5. Use a purpose-built annotation tool to make sure no attribution instance is skipped, since coverage gaps hurt later learning.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

  • I1 (frames): prag could emit an explicit attribution frame for every speech or cognition verb, with source and content slots, even when no discourse connective is present.
  • I2 (world): the world of the text could store reported content as a claim held by its source rather than as a fact, with the speech type deciding how much the wording can be trusted.
  • I3 (tree): after UD parsing, a rule could strip appositive or relative modifiers off the source and adverbial insertions off the cue before the content clause is passed on.
  • I4 (learn): when witness parsers vote on silver attribution frames, agreement could be tracked per slot, so that weak slots like supplement get extra checks or an LLM teacher tier.
  • I5 (learn): a checker could flag every reporting verb in a document that lacks an attribution frame, which would act as a coverage control in evaluation.

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 Semitic languages

 Institutions

🎓Imam Mohammad ibn Saud Islamic UniversityRiyadh · Saudi Arabia6 works · 12 citations

Cite

Cite as Amal Alsaif, Tasniem Alyahya, Madawi Alotaibi, Huda Almuzaini and Abeer Algahtani. 2018. Annotating Attribution Relations in Arabic. Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018).

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

@inproceedings{alsaif2018annotating,
  title = {{Annotating Attribution Relations in Arabic}},
  author = {Amal Alsaif, Tasniem Alyahya, Madawi Alotaibi, Huda Almuzaini and Abeer Algahtani},
  booktitle = {Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018)},
  year = {2018},
  doi = {10.63317/2b3shwzvyczz},
  url = {https://doi.org/10.63317/2b3shwzvyczz}
}

References

50 references, 9 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Al-Saif, Amal, and Katja Markert. 2010. "The Leeds Arabic Discourse Treebank: Annotating Discourse Connectives for Arabic." In LREC. our card
  2. Ali, Mohammed Mohammed Younis. 2004. Introduction to the science of significance and communication (New United Book House). find
  3. Austin, John L. 1975. 'How to do things with words (JO Urmson & M. Sbisa, Eds.)', Harvard U. Press, Cambridge, MA. find
  4. Bouayad, Nawara, and Omar Belkheer. 2012. Categorize speech actions in journalistic discourse Algerian written in Arabic, Journal of Al-Athar Carlson, Lynn, and Daniel Marcu. 2001. Discourse tagging reference manual, ISI Technical Report ISI-TR-545, 54: 56. find
  5. Castor, A. and Pollux, L. E. (1992). The use of user modelling to guide inference and learning. Applied Intelligence, 2(1):37–53. find
  6. Elson, David K, and Kathleen McKeown. 2010. "Automatic Attribution of Quoted Speech in Literary Narrative." In AAAI. Citeseer. find
  7. Elson, David K, and Kathleen R McKeown. 2009. "A tool for deep semantic encoding of narrative texts." In Proceedings of the ACL-IJCNLP 2009 Software Demonstrations, 9-12. Association for Computational Linguistics. find
  8. Fernandes, William Paulo Ducca, Eduardo Motta, and Ruy Luiz Milidiú. 2011. "Quotation extraction for portuguese." In Proceedings of the 8th Brazilian Symposium in Information and Human Language Technology (STIL 2011), Cuiabá, 204-08. find
  9. Grandchercheur, L.B. (1983) Vers une modélisation cognitive de l'être et du néant. In S.G Paris, G.M. Olson, & H.W. Stevenson (Eds.), Fondement des Sciences Cognitives. Hillsdale, NJ: Lawrence Erlbaum Associates, pp. 6-38. find
  10. Guzmán-Cabrera, Rafael, Manuel Montes-y-Gómez, Paolo Rosso, and Luis Villasenor-Pineda. 2009. Using the Web as corpus for self-training text categorization, Information Retrieval, 12: 400-15. find
  11. Huang, Hen-Hsen, and Hsin-Hsi Chen. 2011. "Chinese Discourse Relation Recognition." In IJCNLP, 1442-46. our card
  12. Ide, Nancy, and Keith Suderman. 2009. "Bridging the gaps: interoperability for GrAF, GATE, and UIMA." In Proceedings of the Third Linguistic Annotation Workshop, 27-34. Association for Computational Linguistics. find
  13. Juola, Patrick, and R Harald Baayen. 2005. A controlled-corpus experiment in authorship identification by cross-entropy, Literary and Linguistic Computing, 20: 59-67. find
  14. Kolachina, Sudheer, Rashmi Prasad, Dipti Misra Sharma, and Aravind K Joshi. 2012. "Evaluation of Discourse Relation Annotation in the Hindi Discourse Relation Bank." In LREC, 823-28. our card
  15. Kong, Fang, and Guodong Zhou. 2017. A CDT-styled end-to-end Chinese discourse parser, ACM Transactions on Asian and Low-Resource Language Information Processing (TALLIP), 16: 26. find
  16. Li, Yancui, Wenhe Feng, Jing Sun, Fang Kong, and Guodong Zhou. 2014. "Building Chinese Discourse Corpus with Connective-driven Dependency Tree Structure." EMNLP, 2105-14. find
  17. Maamouri, Mohamed, Ann Bies, Tim Buckwalter, and Wigdan Mekki. 2004. "The penn arabic treebank: Building a large-scale annotated arabic corpus." In NEMLAR conference on Arabic language resources and tools, 466-67. find
  18. Mamede, Nuno, and Pedro Chaleira. 2004. Character identification in children stories. in, Advances in natural language processing (Springer). find
  19. Mírovský, Jiří, Pavlína Jínová, and Lucie Poláková. 2014. "Discourse Relations in the Prague Dependency Treebank 3.0." In Proceedings of COLING 2014, the 25th International Conference on Computational Linguistics: System Demonstrations, 34-38. find
  20. Müller, Christoph, and Michael Strube. 2006. Multi-level annotation of linguistic data with MMAX2, Corpus technology and language pedagogy: New resources, new tools, new methods, 3: 197-214. find
  21. Mustafa, Mansour. 2011. The theory of verbs in the imaginary discourse between Searle and genes, Journal of Al-Athar, 12. find
  22. Neumann, Hendrik, and Martin Schnurrenberger. 2009. E-Mail authorship attribution applied to the Extended Enron Authorship Corpus (XEAC) . find
  23. O'Keefe, Tim, Silvia Pareti, James R Curran, Irena Koprinska, and Matthew Honnibal. 2012. A sequence labelling approach to quote attribution. In Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning, 790-99. Association for Computational Linguistics. find
  24. Ouamour, S, and Halim Sayoud. 2013. "Authorship attribution of short historical arabic texts based on lexical features." In Cyber-Enabled Distributed Computing and Knowledge Discovery (CyberC), 2013 International Conference on, 144-47. IEEE. find
  25. Pareti, Silvia. 2011. Annotating attribution relations and their features. In Proceedings of the fourth workshop on Exploiting Semantic Annotations in Information Retrieval, 19-20. ACM. find
  26. Pareti,Silvia. 2016. PARC 3.0: A Corpus of Attribution Relations, LREC. find
  27. Pareti, Silvia 2012. "A Database of Attribution Relations." In LREC, 3213-17. find
  28. Pareti, Silvia, Timothy O'Keefe, Ioannis Konstas, James R Curran, and Irena Koprinska. 2013. Automatically Detecting and Attributing Indirect Quotations. In EMNLP, 989-99. find
  29. Pareti, Silvia, and Irina Prodanof. 2010. Annotating Attribution Relations: Towards an Italian Discourse Treebank. In LREC. find
  30. Pouliquen, Bruno, Ralf Steinberger, and Clive Best. 2007. Automatic detection of quotations in multilingual news. In Proceedings of Recent Advances in Natural Language Processing, 487-92. find
  31. Prasad, Rashmi, Nikhil Dinesh, Alan Lee, Aravind K Joshi, and Bonnie L Webber. 2006. Attribution and its annotation in the Penn Discourse TreeBank, TAL, 47: 43-64. our card
  32. Prasad, Rashmi, Eleni Miltsakaki, Nikhil Dinesh, Alan Lee, Aravind Joshi, Livio Robaldo, and Bonnie L Webber. 2007. The penn discourse treebank 2.0 annotation manual . our card
  33. Rabab’ah, Abdullateef, Mahmoud Al-Ayyoub, Yaser Jararweh, and Monther Aldwairi. 2016. "Authorship Attribution of Arabic Tweets." In AICCSA Sarmento, Luis, Sergio Nunes, and E Oliveira. 2009. "Automatic extraction of quotes and topics from news feeds." In 4th Doctoral Symposium on Informatics Engineering. find
  34. Saurí, Roser, and James Pustejovsky. 2009. FactBank: a corpus annotated with event factuality, Language Resources and Evaluation, 43: 227. find
  35. Schneider, Nathan, Rebecca Hwa, Philip Gianfortoni, Dipanjan Das, Michael Heilman, Alan Black, Frederick L Crabbe, and Noah A Smith. 2010. Visualizing topical quotations over time to understand news discourse . find
  36. Searle, John R. 1976. A classification of illocutionary acts, Language in society, 5: 1-23. find
  37. Siegel, Sidney, and NJ Castellan. 1988. Nonparametric systems for the behavioural sciences, McGraw Hill International Editions. find
  38. Stenetorp, Pontus, Sampo Pyysalo, Goran Topić, Tomoko Ohta, Sophia Ananiadou, and Jun'ichi Tsujii. 2012. "BRAT: a web-based tool for NLP-assisted text annotation." In Proceedings of the Demonstrations at the 13th Conference of the European Chapter of the Association for Computational Linguistics, 102-07. our card
  39. Strötgen, J. and Gertz, M. (2012). Temporal tagging on different domains: Challenges, strategies, and gold standards. Proceedings of the twelve International Conference on Language Resources and Evaluation (LREC12), pages 3746–3753, Istanbul, Turkey, may. our card
  40. Superman, S., Batman, B., Catwoman, C., and Spiderman, S. (2000). Superheroes experiences with books. The Phantom Editors Associates, Gotham City, 20th edition. find
  41. Wiebe, Janyce. 2002. Instructions for annotating opinions in newspaper articles . find
  42. Wiebe, Janyce, Theresa Wilson, and Claire Cardie. 2005. Annotating expressions of opinions and emotions in language, Language Resources and Evaluation, 39: 165-210. find
  43. Wolf, Florian, and Edward Gibson. 2005. Representing discourse coherence: A corpus-based study, Computational Linguistics, 31: 249-87. our card
  44. Yimam, Seid Muhie, Iryna Gurevych, Richard Eckart de Castilho, and Chris Biemann. 2013. "WebAnno: A Flexible, Web-based and Visually Supported System for Distributed Annotations." In ACL (Conference System Demonstrations), 1-6. our card
  45. Zeyrek, Deniz, and Murathan Kurfalı. 2017. TDB 1.1: Extensions on Turkish Discourse Bank, LAW XI 2017: 76. find
  46. Zhao, Ying, and Justin Zobel. 2005. Effective and scalable authorship attribution using function words. In Asia Information Retrieval Symposium, 174-89. Springer. find
  47. Zhou, Lanjun, Binyang Li, Zhongyu Wei, and Kam-Fai Wong. 2014. "The CUHK Discourse TreeBank for Chinese: Annotating Explicit Discourse Connectives for the Chinese TreeBank." In LREC, 942-49. find
  48. Zhou, Yuping, and Nianwen Xue. 2012. PDTB-style discourse annotation of Chinese text. In Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics: Long Papers-Volume 1, 69-77. Association for Computational Linguistics. find
  49. Yuping Zhou Nianwen Xue. 2015. The chinese discourse treebank: a chinese corpus annotated with discourse relations, Language Resources and Evaluation, 49: 397-431. 10. find
  50. Language Resource References Mohamed Maamouri, Ann Bies, Seth Kulick, Fatma Gaddeche, Wigdan Mekki, Sondos Krouna, Basma Bouziri, Wajdi Zaghouani (2010). The Penn ArabicTreeban: Part 1v 4.1. Distributed via LDC. LDC2010T13. ISLRN 512-715-458-848-0. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/alsaif-2018-annotating-attribution-relations-arabic/