Beyond The Wall Street Journal: Anchoring and Comparing Discourse Signals across Genres

What it is about

The RST Signalling Corpus records which signals mark each RST relation in news, but not where they are. This pilot anchors signals to tokens in 12 GUM documents (11K tokens) from four genres — academic, how-to, interview, news — with the RST-SC taxonomy extended by visual (image) and textual (title, date, attribution) types and genre-layout subtypes; agreement on which tokens signal is kappa 0.77. Of 1,263 signal instances only 16% are discourse markers; 45% are semantic, mostly lexical chains (repeated or related words across the two units), 12% lexical (indicative words such as Tips, Steps, Warnings, or strongly evaluative words), then reference, graphical, syntactic. Markers dominate only condition, concession, antithesis, cause and circumstance; evaluation is never marked by one; sequence is marked by ordinal pairs (first … second), temporal words, and sometimes world knowledge. But is a generic signal shared by contrast, concession and antithesis.

Ideas

• I1. Add a reference/lexical-chain condition to Mova's sentence-order next: keep an unmarked next between consecutive sentences only when the two frames share an entity (same resolved referent, or same lemma) or the second has a temporal or ordinal word; otherwise leave them unlinked. The paper shows chains are the main signal where no connective is present, and adjacency-only next is Mova's weakest role (dev F1 .23). Measure frame bench next F1 and the course «what happened next» questions (the earlier rule-cutting experiment lost course questions, so both measures matter).
• I2. Treat ordinal and sequence word pairs as next signals across sentences: first … then/next/finally, at first … later, numbered steps in how-to lessons. A short closed list in the connective data (see das-2018-constructing-lexicon-english-discourse-connectives I1); measure next precision on lessons from the how-to-like courses.
• I3. Do not try to read but into one relation: map it to the CCR «negative polarity» primitive only (scholman-2016-categories-coherence-relations-discourse-annotation I1) and leave contrast vs concession unspecified unless although/even though or a negated expectation decides it.

For Mova

36. Discourse relations and coherence §6–7 (connectives, implicit relations), §11.1–11.2. Code: prag/src/frames.rs (set_next conditions), the connective table. Measures: frame bench next F1, course test counts on what-next questions.

Topics: Discourse analysis

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.18653/v1/w19-2710

@inproceedings{liu2019beyond,
  title = {{Beyond The Wall Street Journal: Anchoring and Comparing Discourse Signals across Genres}},
  author = {Liu, Yang},
  booktitle = {Proceedings of the Workshop on Discourse Relation Parsing and Treebanking 2019},
  year = {2019},
  doi = {10.18653/v1/w19-2710},
  url = {https://doi.org/10.18653/v1/w19-2710}
}

References

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

1. Amir Zeldes. 2016. rstWeb - A Browser-based Annotation Interface for Rhetorical Structure Theory and Discourse Relations. In Proceedings of NAACL-HLT 2016 System Demonstrations, pages 1–5, San Diego, CA. find
2. Lynn Carlson, Daniel Marcu, and Mary Ellen Okurowski. 2003. Building a Discourse-Tagged Corpus in the Framework of Rhetorical Structure Theory. In Current and New Directions in Discourse and Dialogue, Text, Speech and Language Technology 22, pages 85–112. Kluwer, Dordrecht. find our card
3. Amir Zeldes. 2017. The GUM Corpus: Creating Multilayer Resources in the Classroom. Language Resources and Evaluation, 51(3):581–612. ↗
4. Debopam Das and Maite Taboada. 2017. Signalling of Coherence Relations in Discourse, Beyond Discourse Markers. Discourse Processes, pages 1–29. find
5. Amir Zeldes. 2018a. Multilayer Corpus Studies. Routledge Advances in Corpus Linguistics 22. Routledge, London. find
6. Debopam Das and Maite Taboada. 2018. RST Signalling Corpus: A Corpus of Signals of Coherence Relations. Language Resources and Evaluation, 52(1):149–184. find
7. Amir Zeldes. 2018b. A Neural Approach to Discourse Relation Signaling. Georgetown University Round Table (GURT) 2018: Approaches to Discourse. find
8. Luke Gessler, Yang Liu, and Amir Zeldes. 2019. A Discourse Signal Annotation System for RST Trees. In Proceedings of 7th Workshop on Discourse Relation Parsing and Treebanking (DISRPT) at NAACL-HLT, Minneapolis, MN. (To Appear). find our card
9. Yang Liu and Amir Zeldes. 2019. Discourse relations and signaling information: Anchoring discourse signals in RST-DT. Proceedings of the Society for Computation in Linguistics, 2(1):314–317. find
10. William C Mann and Sandra A Thompson. 1988. Rhetorical Structure Theory: Toward a Functional Theory of Text Organization. Text-Interdisciplinary Journal for the Study of Discourse, 8(3):243–281. find our card
11. Mitchell P. Marcus, Beatrice Santorini, and Mary Ann Marcinkiewicz. 1993. Building a Large Annotated Corpus of English: The Penn Treebank. Special Issue on Using Large Corpora, Computational Linguistics, 19(2):313–330. find our card
12. Rashmi Prasad, Nikhil Dinesh, Alan Lee, Eleni Miltsakaki, Livio Robaldo, Aravind Joshi, and Bonnie Webber. 2008. The Penn Discourse Treebank 2.0. In Proceedings of the 6th International Conference on Language Resources and Evaluation (LREC 2008), pages 2961–2968, Marrakesh, Morocco. find our card
13. Rashmi Prasad, Aravind Joshi, and Bonnie Webber. 2010. Realization of Discourse Relations by Other Means: Alternative Lexicalizations. In Proceedings of the 23rd International Conference on Computational Linguistics: Posters, pages 1023–1031. Association for Computational Linguistics. find
14. Maite Taboada and Debopam Das. 2013. Annotation upon Annotation: Adding Signalling Information to a Corpus of Discourse Relations. D&D, 4(2):249– 281. find
15. Maite Taboada and Julia Lavid. 2003. Rhetorical and Thematic Patterns in Scheduling Dialogues: A Generic Characterization. Functions of Language, 10(2):147–178. find
16. Maite Taboada and William C Mann. 2006. Applications of Rhetorical Structure Theory. Discourse Studies, 8(4):567–588. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/liu-2019-beyond-wall-street-journal-anchoring/