Elamur Research⌕
☰
home / papers / 2015 / Toward Abstractive Summarization Using Semantic Representations

Toward Abstractive Summarization Using Semantic Representations

Summary

The paper investigates whether abstractive summarization can work by manipulating meaning representations instead of copying or compressing source sentences. The authors propose a three-stage pipeline. First, each sentence is parsed into an Abstract Meaning Representation (AMR) graph. Next, these graphs are condensed into one summary graph. Finally, text is produced from that graph. The work concentrates on the middle, graph-to-graph stage.

Sentence graphs are combined into a single connected source graph. Identical concepts are merged, flat named-entity and date fragments are collapsed, and a dummy ROOT node is attached. Edges can optionally be added through expansion. A trainable structured-prediction model then chooses a subgraph of nodes and edges, and the authors compare perceptron, hinge, and ramp loss for learning it.

Experiments use the proxy-report section of the AMR Bank, which supplies gold AMR annotations for both news articles and their summaries. Systems are tested with gold annotations and with JAMR parses. Because no full generator exists yet, a bag of words is read off the summary graph and scored with ROUGE-1.

Ramp loss gives clear gains, and the results suggest that better AMR parsing would improve summaries. This makes the study an early proof of concept for semantics-driven summarization.

Key points

  • Resources: AMR Bank contains 20,341 expert-annotated sentences. JAMR scores 63% smatch F-score on the test data.
  • Dataset: The proxy-report split has 298 train, 35 dev and 33 test documents. Source graphs average about 144 nodes and 221 edges, and expansion grows the edge count roughly 15-fold, to 3,292.
  • Coverage: On test, source graphs cover 63.0% of labeled and 64.7% of unlabeled gold summary edges. Expansion raises coverage to 75.0% at sentence level and 83.3% at document level.
  • Learning: Ramp loss brings substantial gains. Cost-aware hinge loss adds little over perceptron. With JAMR input, ramp loss reaches 51.5% node and 20.0% edge F1 without expansion, and 50.7% and 19.0% with it.
  • Oracle: Oracle ROUGE-1 F1 is 65.8% with gold parses and 57.8% with JAMR parses.

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: Natural language processing

Cited by

108 papers of our library cite it (of the 140884 whose bibliographies we read); 10 of them have a page here, 98 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.3115/v1/n15-1114

@inproceedings{liu2015toward,
  title = {{Toward Abstractive Summarization Using Semantic Representations}},
  author = {Liu, Fei and Flanigan, Jeffrey and Thomson, Sam and Sadeh, Norman and Smith, Noah A.},
  booktitle = {Proceedings of the 2015 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies},
  year = {2015},
  doi = {10.3115/v1/n15-1114},
  url = {https://doi.org/10.3115/v1/n15-1114}
}

References

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

  1. Miguel B. Almeida and Andre F. T. Martins. 2013. Fast and robust compressive summarization with dual decomposition and multi-task learning. In Proceedings of ACL. find
  2. David Bamman and Noah A. Smith. 2013. New alignment methods for discriminative book summarization. In arXiv:1305.1319. — [arXiv:1305.1319](https://arxiv.org/abs/1305.1319) find
  3. Laura Banarescu, Claire Bonial, Shu Cai, Madalina Georgescu, Kira Griffitt, Ulf Hermjakob, Kevin Knight, Philipp Koehn, Martha Palmer, and Nathan Schneider. 2013. Abstract meaning representation for sembanking. In Proceedings of Linguistic Annotation Workshop. find our card
  4. Taylor Berg-Kirkpatrick, Dan Gillick, and Dan Klein. 2011. Jointly learning to extract and compress. In Proceedings of ACL. find
  5. Shu Cai and Kevin Knight. 2013. Smatch: an evaluation metric for semantic feature structures. In Proceedings of ACL. find
  6. Giuseppe Carenini and Jackie Chi Kit Cheung. 2008. Extractive vs. NLG-based abstractive summarization of evaluative text: The effect of corpus controversiality. In Proceedings of the Fifth International Natural Language Generation Conference (INLG). find
  7. Jackie Chi Kit Cheung and Gerald Penn. 2014. Unsupervised sentence enhancement for automatic summarization. In Proceedings of EMNLP. find
  8. Michael Collins. 2002. Discriminative training methods for hidden Markov models: Theory and experiments with perceptron algorithms. In Proceedings of EMNLP. find our card
  9. Hoa Trang Dang and Karolina Owczarzak. 2008. Overview of the TAC 2008 update summarization task. In Proceedings of Text Analysis Conference (TAC). find
  10. Bonnie Dorr, Nizar Habash, and David Traum. 1998. A thematic hierarchy for efficient generation from lexical-conceptual structure. In David Farwell, Laurie Gerber, and Eduard Hovy, editors, Machine Translation and the Information Soup: Proceedings of the Third Conference of the Association for Machi… find
  11. John Duchi, Elad Hazan, and Yoram Singer. 2011. Adaptive subgradient methods for online learning and stochastic optimization. Journal of Machine Learning Research. find
  12. Günes Erkan and Dragomir R. Radev. 2004. LexRank: Graph-based lexical centrality as salience in text summarization. Journal of Artificial Intelligence Research. find
  13. Jeffrey Flanigan, Sam Thomson, Jaime Carbonell, Chris Dyer, and Noah A. Smith. 2014. A discriminative graph-based parser for the abstract meaning representation. In Proceedings of ACL. find
  14. Kavita Ganesan, ChengXiang Zhai, and Jiawei Han. 2010. Opinosis: A graph-based approach to abstractive summarization of highly redundant opinions. In Proceedings of COLING. find our card
  15. Shima Gerani, Yashar Mehdad, Giuseppe Carenini, Raymond T. Ng, and Bita Nejat. 2014. Abstractive summarization of product reviews using discourse structure. In Proceedings of EMNLP. find
  16. Dan Gillick and Benoit Favre. 2009. A scalable global model for summarization. In Proceedings of the NAACL Workshop on Integer Linear Programming for Natural Langauge Processing. find our card
  17. Kevin Gimpel and Noah A. Smith. 2012. Structured ramp loss minimization for machine translation. In Proceedings of NAACL-HLT. find
  18. Zhanying He, Chun Chen, Jiajun Bu, Can Wang, Lijun Zhang, Deng Cai, and Xiaofei He. 2012. Document summarization based on data reconstruction. In Proceedings of AAAI. find
  19. Eduard Hovy, Mitchell Marcus, Martha Palmer, Lance Ramshaw, and Ralph Weischedel. 2006. OntoNotes: The 90% solution. In Proceedings of NAACL. find
  20. Richard M. Karp. 1972. Reducibility Among Combinatorial Problems. In Complexity of Computer Computations, pages 85–103. Springer US. find
  21. Robert T. Kasper. 1989. A flexible interface for linking applications to Penman’s sentence generator. In Proceedings of the DARPA Speech and Natural Language Workshop. find
  22. Gunhee Kim, Leonid Sigal, and Eric P. Xing. 2014. Joint summarization of large-scale collections of web images and videos for storyline reconstruction. In Proceedings of CVPR. find
  23. Kevin Knight, Laura Baranescu, Claire Bonial, Madalina Georgescu, Kira Griffitt, Ulf Hermjakob, Daniel Marcu, Martha Palmer, and Nathan Schneider. 2014. Abstract meaning representation (AMR) annotation release 1.0 LDC2014T12. Web Download. Philadelphia: Linguistic Data Consortium. find
  24. Polina Kuznetsova, Vicente Ordonez, Tamara L. Berg, and Yejin Choi. 2014. TREETALK: Composition and compression of trees for image descriptions. Transactions of ACL. find
  25. Irene Langkilde and Kevin Knight. 1998. Generation that exploits corpus-based statistical knowledge. In Proceedings of COLING. find our card
  26. Chen Li, Yang Liu, Fei Liu, Lin Zhao, and Fuliang Weng. 2014. Improving multi-documents summarization by sentence compression based on expanded constituent parse tree. In Proceedings of EMNLP. find
  27. Hui Lin and Jeff Bilmes. 2010. Multi-document summarization via budgeted maximization of submodular functions. In Proceedings of NAACL. find
  28. Chin-Yew Lin. 2004. ROUGE: a package for automatic evaluation of summaries. In Proceedings of ACL Workshop on Text Summarization Branches Out. find
  29. Fei Liu and Yang Liu. 2013. Towards abstractive speech summarization: Exploring unsupervised and supervised approaches for spoken utterance compression. IEEE Transactions on Audio, Speech, and Language Processing. find
  30. Ivana Ljubić, René Weiskircher, Ulrich Pferschy, Gunnar W. Klau, Petra Mutzel, and Matteo Fischetti. 2006. An Algorithmic Framework for the Exact Solution of the Prize-Collecting Steiner Tree Problem. In Mathematical Progamming, Series B. find
  31. Andre F. T. Martins and Noah A. Smith. 2009. Summarization with a joint model for sentence extraction and compression. In Proceedings of the ACL Workshop on Integer Linear Programming for Natural Language Processing. find
  32. Andre F. T. Martins, Noah A. Smith, and Eric P. Xing. 2009. Concise integer linear programming formulations for dependency parsing. In Proceedings of ACL. find
  33. Ryan McDonald. 2007. A study of global inference algorithms in multi-document summarization. In Proceedings of ECIR. find
  34. Rada Mihalcea and Paul Tarau. 2004. TextRank: Bringing order into text. In Proceedings of EMNLP. find
  35. Courtney Napoles, Benjamin Van Durme, and Chris Callison-Burch. 2011. Evaluating Sentence Compression: Pitfalls and Suggested Remedies. In Proceedings of the Workshop on Monolingual Text-To-Text Generation, MTTG ’11, pages 91–97, Stroudsburg, PA, USA. Association for Computational Linguistics. find
  36. Ani Nenkova and Kathleen McKeown. 2011. Automatic summarization. Foundations and Trends in Information Retrieval. find
  37. Xian Qian and Yang Liu. 2013. Fast joint compression and summarization via graph cuts. In Proceedings of EMNLP. find
  38. Norman Sadeh, Alessandro Acquisti, Travis D. Breaux, Lorrie Faith Cranor, Aleecia M. McDonald, Joel R. Reidenberg, Noah A. Smith, Fei Liu, N. Cameron Russell, Florian Schaub, and Shomir Wilson. 2013. The usable privacy policy project. Technical Report, CMU-ISR-13-119, Carnegie Mellon University. find
  39. Arie Segev. 1987. The Node-Weighted Steiner Tree Problem. Networks, 17(1):1–17. find
  40. Ben Taskar, Carlos Guestrin, and Daphne Koller. 2004. Max-margin Markov networks. In Advances in Neural Information Processing Systems 16. find
  41. Kapil Thadani and Kathleen McKeown. 2013. Sentence compression with joint structural inference. In Proceedings of CoNLL. find
  42. Lucy Vanderwende, Michele Banko, , and Arul Menezes. 2004. Event-centric summary generation. In Proceedings of DUC. find
  43. Bin Zhao and Eric P. Xing. 2014. Quasi real-time summarization for consumer videos. In Proceedings of CVPR. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/liu-2015-abstractive-summarization-semantic-representations/