--- title: "Seeing Things from a Different Angle:Discovering Diverse" description: "Search engines and recommenders rank results by query relevance and popularity. 90% recall. 3 references, cited by 73 papers here." authors: ["Sihao Chen", "Daniel Khashabi", "Wenpeng Yin", "Chris Callison-Burch", "Dan Roth"] year: 2019 venue: "Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)" source: "https://aclanthology.org/N19-1053/" doi: "10.18653/v1/n19-1053" licence: "CC BY 4.0 (ACL Anthology)" licence_url: "https://creativecommons.org/licenses/by/4.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/chen-2019-seeing-things-different-angle-discovering/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/chen-2019-seeing-things-different-angle-discovering/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2019](https://md.research.elamur.ai/free/en/papers/2019/page/4/) / Seeing Things from a Different Angle:Discovering Diverse Perspectives about Claims # Seeing Things from a Different Angle:Discovering Diverse Perspectives about Claims **Authors:** [Sihao Chen](https://md.research.elamur.ai/free/en/authors/sihao-chen/), [Daniel Khashabi](https://md.research.elamur.ai/free/en/authors/daniel-khashabi/), [Wenpeng Yin](https://md.research.elamur.ai/free/en/authors/wenpeng-yin/), [Chris Callison-Burch](https://md.research.elamur.ai/free/en/authors/chris-callison-burch/) and [Dan Roth](https://md.research.elamur.ai/free/en/authors/dan-roth/) · **Year:** 2019 · **Venue:** [Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)](https://md.research.elamur.ai/free/en/venues/naacl/) ([NAACL 2019](https://md.research.elamur.ai/free/en/venues/naacl/2019/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/) **Links:** [Publisher's page](https://aclanthology.org/N19-1053/) · [PDF](https://aclanthology.org/N19-1053.pdf) · [DOI 10.18653/v1/n19-1053](https://doi.org/10.18653/v1/n19-1053) · [OpenAlex](https://openalex.org/W2946364534) **Citations:** 99 citations (OpenAlex, 2026-10-14) · [73](https://md.research.elamur.ai/free/en/papers/chen-2019-seeing-things-different-angle-discovering/cited-by/) in our library, of the 141,840 papers whose bibliographies we read ## Summary Search engines and recommenders rank results by query relevance and popularity. They do not check whether the viewpoints shown are varied or backed by evidence, and this skews how people encounter contested claims. The authors frame this as a language understanding problem called substantiated perspective discovery. Given a claim, a system must pick out a compact, non-redundant set of perspectives, label each one's stance, and attach evidence paragraphs that support it. To study the task they built PERSPECTRUM. It started from material crawled from debate websites, was enlarged with search-engine results and paraphrases, and was cleaned through several crowdsourcing rounds that dropped items with low annotator agreement. The problem is split into subtasks: finding relevant perspectives, classifying stance, judging whether two perspectives are equivalent with respect to a claim, and retrieving supporting evidence. The authors evaluated IR and BERT baselines alongside human performance. BERT clearly beats IR, but humans still lead by a wide margin. The resource therefore gives the NLP community an open benchmark for work on reducing selection bias. ## Key points - PERSPECTRUM contains 907 claims, 11,164 perspectives and 8,092 evidence paragraphs. The seed data came from idebate.com, debatewise.org and procon.org, and the dataset is released at github.com/CogComp/perspectrum. - Perspective equivalence is defined relative to a specific claim. Two perspectives may count as equivalent for one claim and not for another, which is what separates this subtask from ordinary paraphrase detection. - On perspective equivalence, BERT improves on the IR baseline by about 36%, yet it still falls roughly 20% short of human performance. - For evidence extraction, long paragraphs are split into sliding windows of 3 sentences for BERT. A paragraph counts as supporting when enough of its windows are predicted positive, with that share tuned on the dev set. Scores on this subtask are lower than on the others. - An IR system generates candidates for the other models. Its top-15 perspectives reach over 90% recall, and its top-60 evidence candidates reach over 85% recall. The authors leave detecting argue-worthy claims, as well as source trustworthiness and credibility, to future work. *Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.* ## Institutions - [University of Pennsylvania](https://md.research.elamur.ai/free/en/universities/university-of-pennsylvania/) — University, Philadelphia, United States · 170 works · 60,739 citations 📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 9,086 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 8,889 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 9,296 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 9,206 works ## Cite Cite as: Sihao Chen, Daniel Khashabi, Wenpeng Yin, Chris Callison-Burch and Dan Roth. 2019. Seeing Things from a Different Angle:Discovering Diverse Perspectives about Claims. Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{chen2019seeing, title = {{Seeing Things from a Different Angle:Discovering Diverse Perspectives about Claims}}, author = {Sihao Chen, Daniel Khashabi, Wenpeng Yin, Chris Callison-Burch and Dan Roth}, booktitle = {Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)}, year = {2019}, doi = {10.18653/v1/n19-1053}, url = {https://doi.org/10.18653/v1/n19-1053} } ``` - [← previous Multilingual Neural Machine Translation with Language Clustering](https://md.research.elamur.ai/free/en/papers/tan-2019-multilingual-neural-machine-translation-language/) - [next → Soft Contextual Data Augmentation for Neural Machine Translation](https://md.research.elamur.ai/free/en/papers/gao-2019-soft-contextual-data-augmentation-neural/) - [↑ back to papers of 2019](https://md.research.elamur.ai/free/en/papers/2019/page/4/) ## References 3 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. Ehud Aharoni, Anatoly Polnarov, Tamar Lavee, Daniel Hershcovich, Ran Levy, Ruty Rinott, Dan Gutfreund, and Noam Slonim. 2014. [A benchmark dataset for automatic detection of claims and evidence in the context of controversial topics](https://md.research.elamur.ai/free/en/papers/aharoni-2014-benchmark-dataset-automatic-detection-claims/). In Proceedings of the First Workshop on Argumentation Mining, pages… 2. Yamen Ajjour, Wei-Fan Chen, Johannes Kiesel, Henning Wachsmuth, and Benno Stein. 2017. Unit Segmentation of Argumentative Texts. In Proceedings of the Fourth Workshop on Argument Mining (ArgMining 2017), pages 118–128. [find](https://scholar.google.com/scholar?q=Yamen+Ajjour%2C+Wei%2DFan+Chen%2C+Johannes+Kiesel%2C+Henning+Wachsmuth%2C+and+Benno+Stein%2E+2017%2E+Unit+Segmentation+of+Argumentative+Texts%2E+In+Proceedings+of+the+Fourth+Workshop+on+Argument+Mining+%28ArgMining+201) 3. Tariq Alhindi, Savvas Petridis, Smar Muresan, and a. 2018. [Where is your Evidence: Improving Fact-checking by Justification Modeling](https://md.research.elamur.ai/free/en/papers/alhindi-2018-where-is-your-evidence-improving/). In Proceedings of the First Workshop on Fact Extraction and VERification (FEVER), pages 85–90. \#\# Conclusion *Updated 24 October 2026*