Elamur Research⌕
☰
home / papers / 2018 / Identifying the Most Dominant Event in a News Article by Mining Event Coreference Relations

Identifying the Most Dominant Event in a News Article by Mining Event Coreference Relations

Summary

This short paper introduces a new task: finding the single event that dominates a news article. This central event ties together the story's foreground and background events, and it is not necessarily the event in the headline or opening sentence. The authors annotated central events in 104 news articles from the RED and KBP 2015 corpora. They observed that central events tend to be mentioned many times and spread across the whole text. They also tend to appear early, have many sub-events, and are usually actual rather than hypothetical or generic.

From these observations they built rule-based classifiers over gold event coreference chains. The rules rank events by:

  • chain size
  • stretch
  • first-mention position
  • sub-event count
  • realis status

The best rule system clearly beats headline/first-sentence heuristics, PageRank-style random walks and regression models. However, it drops to near-baseline accuracy when using predicted coreference chains. The work matters for summarization, storyline generation and text segmentation. It also shows that better event coreference resolution is a bottleneck for these applications.

Key points

  • New resource: central-event annotations for 30 RED and 74 KBP 2015 news articles. Two annotators chose the same event in 97 of 104 documents (93%).
  • In roughly 75% of documents, the central event has either the most coreferential mentions or the widest sentence span in its coreference chain.
  • The best system, Coreference + Subevent + Realis, reaches F1 72.73 on RED and 72.05 on KBP 2015. Adding sub-event and realis constraints raises precision by over 11 points compared with coreference rules alone.
  • Baselines underperform the rules: heuristics and random-walk ranking score F1 ≤ 45.94, and linear regression and SVR score about 63–67.
  • With system-predicted coreference, F1 falls to 45.0 (RED) and 47.05 (KBP 2015). Both missed and spurious links distort the size and stretch ranks.

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

18 papers of our library cite it (of the 140673 whose bibliographies we read); 4 of them have a page here, 14 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/n18-2055

@inproceedings{choubey2018identifying,
  title = {{Identifying the Most Dominant Event in a News Article by Mining Event Coreference Relations}},
  author = {Choubey, Prafulla Kumar and Raju, Kaushik and Huang, Ruihong},
  booktitle = {Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (Short Papers)},
  year = {2018},
  doi = {10.18653/v1/n18-2055},
  url = {https://doi.org/10.18653/v1/n18-2055}
}

References

22 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Sean Monahan and Mary Brunson. 2014. Qualities of eventiveness. In Proceedings of the Second Workshop on EVENTS: Definition, Detection, Coreference, and Representation. pages 59–67. find
  2. Sergey Brin and Lawrence Page. 2012. Reprint of: The anatomy of a large-scale hypertextual web search engine. Computer networks 56(18):3825–3833. find
  3. Tim O’Gorman, Kristin Wright-Bettner, and Martha Palmer. 2016. Richer event description: Integrating event coreference with temporal, causal and bridging annotation. Computing News Storylines page 47. find our card
  4. François Chollet. 2015. Keras. https://github.com/fchollet/keras. ↗
  5. Nan Decker. 1985. The use of syntactic clues in discourse processing. In Proceedings of the 23rd annual meeting on Association for Computational Linguistics. Association for Computational Linguistics, pages 315–323. find
  6. F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and E. Duchesnay. 2011. Scikit-learn: Machine learning in Python. Journal of Machine Learning Research 12:2825… find
  7. Joe Ellis, Jeremy Getman, Dana Fore, Neil Kuster, Zhiyi Song, Ann Bies, and Stephanie Strassel. 2015. Overview of linguistic resources for the tac kbp 2015 evaluations: Methodologies and results. In Proceedings of TAC KBP 2015 Workshop, National Institute of Standards and Technology. pages 16–17. find
  8. Jeffrey Pennington, Richard Socher, and Christopher D Manning. 2014. Glove: Global vectors for word representation. In EMNLP. volume 14, pages 1532– 1543. find
  9. Elena Filatova and Vasileios Hatzivassiloglou. 2004a. Event-based extractive summarization. In Proceedings of ACL Workshop on Summarization. Barcelona, Spain., volume 111. find
  10. Tsunoda Tasaku. 1981. Split case-marking patterns in verb-types and tense/aspect/mood. Linguistics 19(5-6):389–438. find
  11. Elena Filatova and Vasileios Hatzivassiloglou. 2004b. A formal model for information selection in multi-sentence text extraction. In Proceedings of the 20th international conference on Computational Linguistics. Association for Computational Linguistics, page 397. find
  12. K. Toutanova, D. Klein, C. Manning, and Y. Singer. 2003. Feature-Rich Part-of-Speech Tagging with a Cyclic Dependency Network. In Proceedings of HLT-NAACL 2003. find
  13. Tasaku Tsunoda. 1985. Remarks on transitivity. Journal of linguistics pages 385–396. find
  14. Francis Galton. 1886. Regression towards mediocrity in hereditary stature. The Journal of the Anthropological Institute of Great Britain and Ireland 15:246–263. find
  15. Shyam Upadhyay, Christos Christodoulopoulos, and Dan Roth. 2016. ” making the news”: Identifying noteworthy events in news articles. In Proceedings of the Fourth Workshop on Events. pages 1–7. find
  16. Joseph Evans Grimes. 1975. The thread of discourse, volume 207. Walter de Gruyter. find
  17. Vladimir N. Vapnik. 1995. The Nature of Statistical Learning Theory. Springer-Verlag New York, Inc., New York, NY, USA. find
  18. Paul J Hopper and Sandra A Thompson. 1980. Transitivity in grammar and discourse. language pages 251–299. find
  19. Roderick Kay and Ruth Aylett. 1996. Transitivity and foregrounding in news articles: Experiments in information retrieval and automatic summarising. In Proceedings of the 34th Annual Meeting on Association for Computational Linguistics. Association for Computational Linguistics, Stroudsburg, PA, USA… find
  20. Mike Lewis, Luheng He, and Luke Zettlemoyer. 2015. Joint a* ccg parsing and semantic role labelling. In EMNLP. pages 1444–1454. find
  21. Rada Mihalcea and Paul Tarau. 2004. Textrank: Bringing order into text. In EMNLP. volume 4, pages 404–411. find
  22. Teruko Mitamura, Zhengzhong Liu, and Eduard Hovy. 2015. Overview of tac kbp 2015 event nugget track. In Text Analysis Conference. find

↑ back to top

This page in Markdown →