Elamur Research⌕
☰
home / papers / 2023 / ISO 24617-2 on a cusp of languages

ISO 24617-2 on a cusp of languages

Paper · 2023 · Krzysztof Hwaszcz, Marcin Oleksy, Aleksandra Domogała et al.

Summary

The paper asks how well ISO 24617-2, a dialogue act annotation standard meant to work for any language but illustrated mostly with English, fits Polish. The authors use their experience building DiaBiz.Kom, an annotation layer on 1100 call-center transcripts from the DiaBiz corpus. They compare Polish and English on several points: honorific address forms, inflection taking over the job of word order, freer question word order, a higher share of complex sentences, and culture-specific conversational habits. Polish call openings, for example, tend to be one-directional and do not follow an offer-then-accept pattern. Polysemous high-frequency words caused the most trouble, especially "proszę" and "tak", which can carry many different communicative functions. The team handled this with guidelines, substitution tests, and repeated rounds of review. Their conclusion is that the standard's existing categories are enough to cover Polish, but only if context is weighed carefully and the guidelines rest on a clearer theoretical basis with examples from more languages than English.

Key points

  • DiaBiz contains 3,766 Polish telephone calls from business settings, built on 110 scripts. DiaBiz.Kom annotates 1100 of these transcripts, 10 per script, following ISO 24617-2:2012, with the 2019 edition also consulted.
  • Each dialogue goes through three people: two annotators working independently, then a super-annotator who resolves their disagreements. Communicative functions and dimensions were annotated first, and functional and dependence relations were added afterwards.
  • The corpus has 1,277,965 tokens and 151,520 final communicative-function annotations. A sample is available under CC BY-NC-ND 4.0.
  • The Polish word "tak" received four function labels: Confirm (739), Contact Indication (1311), Auto/Allo Positive (257) and Agreement (326). "Proszę" was labeled with functions including acceptOffer, autoNegative, contactIndication and contactCheck.
  • A cited study found complex sentences make up 26.23% of formal Polish text versus 18.97% of English. Polish "Can I help you?"-style openings were labeled Interaction-Structuring rather than Offer.

Interesting ideas

  • I1. Each item gets two independent annotations, and a super-annotator settles the conflicts, so disagreements surface and are resolved in one place rather than averaged away.
  • I2. The function of a short reply such as "yes" depends on the act before it: it is Confirm after a checkQuestion and Agreement after an assertion, while a rising "Yes?" signals Contact Indication.
  • I3. Substitution tests decide a polysemous word's function by checking which unambiguous paraphrase fits ("excuse me?" versus "you're welcome" versus "go on").
  • I4. When a routine exchange has a different shape in another language, the authors relabel it by its real role in the conversation (Interaction-Structuring) rather than forcing the English Offer/accept template onto it.
  • I5. The team took a structural stance on polysemy: same-spelled items with clearly different meanings are treated as separate lexical units, which makes annotation decisions more crisp.

For Mova

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

  • I1 (learn): Silver data from several witness parsers could be routed the same way, with an adjudicator step (an LLM teacher tier or a rule) that resolves only the disputed cases and logs why.
  • I2 (frames): The speech-act labeler in prag could take the previous turn's act as a key feature for tagging replies like "yes", "okay" and "please", instead of reading them from lexical form alone.
  • I3 (frames): Each discourse-marker sense could carry a canonical paraphrase, so every sense choice is explainable as "this substitution fits the context."
  • I4 (back): Translation through the meta-world should map dialogue openings and politeness formulas by function, producing whatever form the target language uses for that role.
  • I5 (tokens): Lemma entries for multifunction words could be split into distinct sense-units, so later stages receive a discrete choice instead of one blurred lemma.

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, topics and concepts

📚 Shelves 2The field of science this work belongs to (OpenAlex fields).Arts and Humanities · 17,450 worksLanguage and Linguistics · 17,296 works

🧭 Topics 1Research subjects and languages the work is about.English · 2,379 works

Cite

Cite as Krzysztof Hwaszcz, Marcin Oleksy, Aleksandra Domogała and Jan Wieczorek. 2023. ISO 24617-2 on a cusp of languages. Proceedings of the 19th Joint ACL-ISO Workshop on Interoperable Semantics (ISA-19).

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

@inproceedings{hwaszcz2023iso,
  title = {{ISO 24617-2 on a cusp of languages}},
  author = {Krzysztof Hwaszcz, Marcin Oleksy, Aleksandra Domogała and Jan Wieczorek},
  booktitle = {Proceedings of the 19th Joint ACL-ISO Workshop on Interoperable Semantics (ISA-19)},
  year = {2023},
  url = {https://aclanthology.org/2023.isa-1.6/}
}

References

24 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Chengyu Fang, Jing Cao, Harry Bunt, and Xiaoyue Liu. 2012. The annotation of the switchboard corpus with the new iso standard for dialogue act analysis. In Proceedings of the Eighth Joint ACL-ISO Workshop on Interoperable Semantic Annotation. Eighth Joint ACL-ISO Workshop on Interoperable Semantic Annotation ; Conference date: 03-10-2012 Through 05-10-2012. find
  2. Ju D Apresjan. 1974. Leksičeskaja semantika: Sinonim-ičeskije sredstva jazyka. Nauka. find
  3. Andrzej Bogusławski. 1976. O zasadach rejestracji jednostek j˛ezyka. Poradnik j˛ezykowy, 8(342):356– 364. find
  4. Harry Bunt, Jan Alexandersson, Jean Carletta, Jae-Woong Choe, Alex Chengyu Fang, Koiti Hasida, Kiyong Lee, Volha Petukhova, Andrei Popescu-Belis, Laurent Romary, Claudia Soria, and David Traum. 2010a. Towards an ISO Standard for Dialogue Act Annotation. In Seventh conference on International Language Resources and Evaluation (LREC’10), La Valette, Malta.
  5. Emer Gilmartin, Christian Saam, Brendan Spillane, Maria O’Reilly, Ketong Su, Arturo Calvo, Loredana Cerrato, Killian Levacher, Nick Campbell, and Vincent Wade. 2018. The ADELE corpus of dyadic social text conversations:dialog act annotation with ISO 24617-2. In Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018), Miyazaki, Japan. European Language Resources Association (ELRA).
  6. Harry Bunt, Jan Alexandersson, Jean Carletta, Jae-Woong Choe, Alex Chengyu Fang, Koiti Hasida, Kiyong Lee, Volha Petukhova, Andrei Popescu-Belis, Laurent Romary, Claudia Soria, and David Traum. 2010b. Towards an ISO standard for dialogue act annotation. In Proceedings of the Seventh International Conference on Language Resources and Evaluation (LREC’10), Valletta, Malta. European Language Resources Association (ELRA).
  7. Ewa Gruszczyńska, Małgorzata Guławska-Gawkowska, and Anna Szcz˛esny. 2019. Translatoryczne i dyskursywne oblicza komunikacji. Instytutu Lingwistyki Stosowanej WLS UW. find
  8. Simon Keizer, Harry Bunt, and Volha Petukhova. 2011. Multidimensional Dialogue Management, pages 57– 86. Springer Berlin Heidelberg, Berlin, Heidelberg. find
  9. George Lakoff. 2008. Women, fire, and dangerous things: What categories reveal about the mind. University of Chicago press. find
  10. Harry Bunt, Jan Alexandersson, Jae-Woong Choe, Alex Chengyu Fang, Koiti Hasida, Volha Petukhova, Andrei Popescu-Belis, and David Traum. 2012. ISO 24617-2: A semantically-based standard for dialogue annotation. In Proceedings of the Eighth International Conference on Language Resources and Evaluation (LREC’12), pages 430–437, Istanbul, Turkey. European Language Resources Association (ELRA).
  11. Barbara Lewandowska-Tomaszczyk and Marcel Thelen. 2013. Translation and meaning part 10. find
  12. Essa T. Mohammed. 2009. Polysemy as a lexical problem in translation. find
  13. Harry Bunt, Volha Petukhova, Emer Gilmartin, Catherine Pelachaud, Alex Chengyu Fang, Simon Keizer, and Laurent Prévot. 2020. The iso standard for dialogue act annotation, second edition. In International Conference on Language Resources and Evaluation.
  14. Thi-Lan Ngo, Son-Bao Pham, Khac-Linh Pham, Xuan-Hieu Phan, and Minh-Son Cao. 2017. Dialogue act segmentation for vietnamese human-human conversational texts. In 2017 9th International Conference on Knowledge and Systems Engineering (KSE), pages 203–208. find
  15. Harry Bunt, Volha Petukhova, Andrei Malchanau, Kars Wijnhoven, and Alex Fang. 2016. The DialogBank. In Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC’16), pages 3151–3158, Portorož, Slovenia. European Language Resources Association (ELRA).
  16. Marcin Oleksy, Jan Wieczorek, Dorota Drużyłowska, Julia Klyus, Aleksandra Domogała, Krzysztof Hwaszcz, Hanna K˛edzierska, Daria Mikoś, and Anita Wróż. 2022. Diabiz.kom - towards a polish dialogue act corpus based on iso 24617-2 standard. In International Conference on Computational Linguistics. find
  17. Harry Bunt, James Pustejovsky, and Kiyong Lee. 2018. Towards an ISO standard for the annotation of quantification. In Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018), Miyazaki, Japan. European Language Resources Association (ELRA). find
  18. Mateusz Arkadiusz Ostalak. 2019. A comparative analysis of grammatical structures and vocabulary in Polish and English Facebook chats. Katowice. find
  19. Volha Petukhova, Harry Bunt, Andrei Malchanau, and Ramkumar Aruchamy. 2015. Experimenting with grounding strategies in dialogue. SEMDIAL 2015 goDIAL, page 198. find
  20. Shammur Absar Chowdhury, Evgeny Stepanov, and Giuseppe Riccardi. 2016. Transfer of corpus-specific dialogue act annotation to ISO standard: Is it worth it? In Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC’16), pages 132–135, Portorož, Slovenia. European Language Resources Association (ELRA).
  21. Volha Petukhova, Martin Gropp, Dietrich Klakow, Gregor Eigner, Mario Topf, Stefan Srb, Petr Motlicek, Blaise Potard, John Dines, Olivier Deroo, Ronny Egeler, Uwe Meinz, Steffen Liersch, and Anna Schmidt. 2014. The DBOX corpus collection of spoken human-human and human-machine dialogues. In Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC’14), pages 252–258, Reykjavik, Iceland. European Language Resources Association (ELRA). find
  22. Piotr P˛ezik, Gosia Krawentek, Sylwia Karasińska, Paweł Wilk, Paulina Rybińska, Anna Cichosz, Angelika Peljak-Lapinska, Mikołaj Deckert, and Michał Adamczyk. 2022. Diabiz – an annotated corpus of polish call center dialogs. In International Conference on Language Resources and Evaluation. find
  23. Goran Schmidt. 2008. Polysemy in translation - selecting the right sense. find
  24. Purificação Silvano, Mariana Damova, Giedrė Valūnaitė Olekeviienė, Chaya Liebeskind, C. Chiarcos, Dimitar Trajanov, Ciprian-Octavian Truică, Elena Simona Apostol, and Anna B aczkowska. 2022. Iso-based annotated multilingual parallel corpus for discourse markers. In International Conference on Language Resources and Evaluation. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/hwaszcz-2023-iso-24617-2-cusp-languages/