--- title: "The BioScope corpus — Szarvas 2008" description: "The authors tackle a gap in biomedical text mining: systems must tell factual statements apart from negated or speculative ones, but no sizeable public…" authors: ["György Szarvas", "Veronika Vincze", "Richárd Farkas", "János Csirik"] year: 2008 venue: "Proceedings of the Workshop on Current Trends in Biomedical Natural Language Processing" source: "https://aclanthology.org/W08-0606/" licence: "CC BY-NC-SA 3.0 (ACL Anthology, before 2016)" licence_url: "https://creativecommons.org/licenses/by-nc-sa/3.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/szarvas-2008-bioscope-corpus-annotation-negation-uncertainty/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/szarvas-2008-bioscope-corpus-annotation-negation-uncertainty/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2008](https://md.research.elamur.ai/free/en/papers/2008/) / The BioScope corpus: annotation for negation, uncertainty and their scope in biomedical texts # The BioScope corpus: annotation for negation, uncertainty and their scope in biomedical texts **Authors:** [György Szarvas](https://md.research.elamur.ai/free/en/authors/gyorgy-szarvas/), [Veronika Vincze](https://md.research.elamur.ai/free/en/authors/veronika-vincze/), [Richárd Farkas](https://md.research.elamur.ai/free/en/authors/richard-farkas/), [János Csirik](https://md.research.elamur.ai/free/en/authors/janos-csirik/) · **Year:** 2008 · **Venue:** Proceedings of the Workshop on Current Trends in Biomedical Natural Language Processing · **Licence:** [CC BY-NC-SA 3.0 (ACL Anthology, before 2016)](https://creativecommons.org/licenses/by-nc-sa/3.0/) **Links:** [Publisher's page](https://aclanthology.org/W08-0606/) · [PDF](https://aclanthology.org/W08-0606.pdf) · [OpenAlex](https://openalex.org/W2043335066) **Citations:** 171 citations (openalex-snapshot · 2026-10-14) · [69](#cited-by) in our library, of the 140673 papers whose bibliographies we read ## Summary The authors tackle a gap in biomedical text mining: systems must tell factual statements apart from negated or speculative ones, but no sizeable public corpus existed for training and evaluating negation and hedge detection or for resolving the scope of those cues. They built BioScope, a freely available corpus of more than 20,000 sentences in three parts: clinical (radiology-style) free texts, full biological papers and biological abstracts. Negative and speculative keywords are marked at the token level, and their linguistic scope is marked within the sentence. Two linguists annotated the texts independently, and a chief annotator, who also wrote the guidelines, resolved their disagreements. Agreement is reported as F-measure for keywords and for left, right and full scope boundaries. The corpus shows that clinical language is relatively easy: cue words there almost always function as cues. Scientific text is much harder because candidate cues are highly ambiguous and sentences are long. Hedging proved harder than negation for both machines and humans. The resource allows statistical systems to be compared on common ground. ## Key points - Subcorpus sizes: clinical texts have 1954 documents and 6383 sentences; full papers have 9 documents and 2624 sentences; abstracts have 1273 documents and 11872 sentences. In total there are over 20,000 sentences. - The full-paper part includes the Hedge classification corpus of Medlock and Briscoe (2007), which originally had only sentence-level hedge labels. BioScope enriches it with negation cues and scope annotation for both phenomena. - Ambiguity differs sharply by domain. The word "or" acts as a speculative cue in only 11.32% of its occurrences in abstracts, compared with 97.86% in clinical texts. Some hedge cue candidates are non-cues in up to about 90% of cases in abstracts. - Across the corpus, over 20% of sentences contain a negation or hedge that substantially changes their meaning. - Related resources such as Genia Event and BioInfer label negation and uncertainty on events and relations, but they do not mark the triggering keywords. BioScope presents itself as the first corpus with both cues and their scopes. ## Interesting ideas - **Min-max annotation strategy.** Cues are marked as the smallest unit that expresses negation or hedging, while scopes are stretched to the largest syntactic unit. Clear rules like this at both ends reduce annotator hesitation and could be applied to other cue-plus-span tasks, such as evidentiality or sentiment triggers. - **Complex keywords defined by compositionality.** A multiword phrase counts as one cue only if its parts cannot express the hedge or negation on their own. Otherwise each word is annotated separately and the scopes nest. This test is reusable for multiword expressions in any language. - **Annotate by meaning, not surface form.** Grammatically negative words are not labelled when the sentence is really a speculation or a plain assertion. Questions are not treated as hedges by default. This prevents naive keyword matching from defining the gold standard. - **Fine-grained agreement metrics.** Agreement is measured as F-measure with one annotator treated as gold, and computed separately for left boundary, right boundary and exact full scope. This shows where annotators actually diverge and suits any span annotation project. - **Per-keyword ambiguity ratios per domain.** The authors report how often each candidate word is a real cue in each domain. This quickly shows why rule-based approaches succeed in one genre and fail in another, which is useful when moving a pipeline to new text types or languages. *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](https://md.research.elamur.ai/free/en/shelves/linguistics/) · [Natural language processing](https://md.research.elamur.ai/free/en/shelves/nlp/) ## Cited by Cited by 69 papers of our library cite it (of the 140673 whose bibliographies we read); 4 of them have a page here, 65 not yet. - [Modality and Negation in SIMT Use of Modality and Negation in Semantically-Informed Syntactic MT](https://md.research.elamur.ai/free/en/papers/baker-2012-modality-negation-simt-use-modality/) - [A Lexicon-Based Approach for Detecting Hedges in Informal Text](https://md.research.elamur.ai/free/en/papers/islam-2020-lexicon-based-approach-detecting-hedges/) - [Did It Happen? The Pragmatic Complexity of Veridicality Assessment](https://md.research.elamur.ai/free/en/papers/marneffe-2012-did-it-happen-pragmatic-complexity/) - [Learning the Scope of Hedge Cues in Biomedical Texts](https://md.research.elamur.ai/free/en/papers/morante-2009-learning-scope-hedge-cues-biomedical/) ## Cite The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{szarvas2008bioscope, title = {{The BioScope corpus: annotation for negation, uncertainty and their scope in biomedical texts}}, author = {Szarvas, György and Vincze, Veronika and Farkas, Richárd and Csirik, János}, booktitle = {Proceedings of the Workshop on Current Trends in Biomedical Natural Language Processing}, year = {2008}, doi = {10.3115/1572306.1572314}, url = {https://doi.org/10.3115/1572306.1572314} } ``` - [← previous The Annotation Guidelines of the Latin Dependency Treebank and Index Thomisticus Treebank: the Treatment of some specific Syntactic Constructions in Latin](https://md.research.elamur.ai/free/en/papers/bamman-2008-annotation-guidelines-latin-dependency-treebank/) - [next → The CoNLL 2008 Shared Task on Joint Parsing of Syntactic and Semantic Dependencies](https://md.research.elamur.ai/free/en/papers/surdeanu-2008-conll-2008-shared-task-joint/) - [↑ back to papers of 2008](https://md.research.elamur.ai/free/en/papers/2008/) ## References 14 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. Wendy W. Chapman, Will Bridewell, Paul Hanbury, Gregory F. Cooper and Bruce G. Buchanan. 2001. A Simple Algorithm for Identifying Negated Findings and Diseases in Discharge Summaries. Journal of Biomedical Informatics, 34(5):301–310. [↗](https://doi.org/10.1006/jbin.2001.1029) 2. N. Collier, H. S. Park, N. Ogata, Y. Tateishi, C. Nobata, T. Ohta, T. Sekimizu, H. Imai, K. Ibushi, and J. Tsujii. 1999. The GENIA project: corpus-based knowledge acquisition and information extraction from genome research papers. Proceedings of EACL-99. [find](https://scholar.google.com/scholar?q=N%2E+Collier%2C+H%2E+S%2E+Park%2C+N%2E+Ogata%2C+Y%2E+Tateishi%2C+C%2E+Nobata%2C+T%2E+Ohta%2C+T%2E+Sekimizu%2C+H%2E+Imai%2C+K%2E+Ibushi%2C+and+J%2E+Tsujii%2E+1999%2E+The+GENIA+project%3A+corpus%2Dbased+knowledge+acquisition+and+information+extractio) 3. Peter L. Elkin, Steven H. Brown, Brent A. Bauer, Casey S. Husser, William Carruth, Larry R. Bergstrom and Dietlind L. Wahner-Roedler. 2005. A controlled trial of automated classification of negation from clinical notes. BMC Medical Informatics and Decision Making 5:13 doi:10.1186/1472-6947-5-13. [↗](https://doi.org/10.1186/1472-6947-5-13) 4. C. Friedman, P.O. Alderson, J.H. Austin, J.J. Cimino, and S.B. Johnson. 1994. A general natural-language text processor for clinical radiology. Journal of the American Medical Informatics Association, 1(2):161–174. [↗](https://doi.org/10.1136/jamia.1994.95236146) 5. Yang Huang and Henry J. Lowe. 2007. A Novel Hybrid Approach to Automated Negation Detection in Clinical Radiology Reports. Journal of the American Medical Informatics Association, 14(3):304–311. [↗](https://doi.org/10.1197/jamia.m2284) 6. Ken Hyland. 1994. Hedging in academic writing and EAP textbooks. English for Specific Purposes, 13(3):239–256. [find](https://scholar.google.com/scholar?q=Ken+Hyland%2E+1994%2E+Hedging+in+academic+writing+and+EAP+textbooks%2E+English+for+Specific+Purposes%2C+13%283%29%3A239%E2%80%93256%2E) 7. Jin-Dong Kim, Tomoko Ohta, and Jun'ichi Tsujii. 2008. Corpus annotation for mining biomedical events from literature. [↗](https://doi.org/10.1186/1471-2105-9-10) 8. BMC Bioinformatics 2008, 9:10. [find](https://scholar.google.com/scholar?q=BMC+Bioinformatics+2008%2C+9%3A10%2E) 9. Marc Light, Xin Ting Qui, and Padmini Srinivasan. 2004. The language of bioscience: Facts, speculations, and statements in between. In Proceedings of BioLink 2004 Workshop on Linking Biological Literature, Ontologies and Databases: Tools for Users. Boston, Massachusetts, Association for Computational Linguistics, 17–24. [find](https://scholar.google.com/scholar?q=Marc+Light%2C+Xin+Ting+Qui%2C+and+Padmini+Srinivasan%2E+2004%2E+The+language+of+bioscience%3A+Facts%2C+speculations%2C+and+statements+in+between%2E+In+Proceedings+of+BioLink+2004+Workshop+on+Linking+Biological+Litera) 10. Ben Medlock and Ted Briscoe. 2007. Weakly supervised learning for hedge classification in scientific literature. In Proceedings of the 45th Annual Meeting of the Association of Computational Linguistics. Prague, Association for Computational Linguistics, 992– 999. [find](https://scholar.google.com/scholar?q=Ben+Medlock+and+Ted+Briscoe%2E+2007%2E+Weakly+supervised+learning+for+hedge+classification+in+scientific+literature%2E+In+Proceedings+of+the+45th+Annual+Meeting+of+the+Association+of+Computational+Linguisti) [our card](https://md.research.elamur.ai/free/en/papers/medlock-2007-weakly-supervised-learning-hedge-classification/) 11. Pradeep G. Mutalik, Aniruddha Deshpande, and Prakash M. Nadkarni. 2001. Use of General-purpose Negation Detection to Augment Concept Indexing of Medical Documents: A Quantitative Study Using the UMLS. Journal of the American Medical Informatics Association, 8(6):598–609. [↗](https://doi.org/10.1136/jamia.2001.0080598) 12. John P. Pestian, Chris Brew, Pawel Matykiewicz, DJ Hovermale, Neil Johnson, K. Bretonnel Cohen, and Wlodzislaw Duch. 2007. A shared task involving multi-label classification of clinical free text. In Biological, translational, and clinical language processing. Prague, Association for Computational Linguistics, 97–104. [find](https://scholar.google.com/scholar?q=John+P%2E+Pestian%2C+Chris+Brew%2C+Pawel+Matykiewicz%2C+DJ+Hovermale%2C+Neil+Johnson%2C+K%2E+Bretonnel+Cohen%2C+and+Wlodzislaw+Duch%2E+2007%2E+A+shared+task+involving+multi%2Dlabel+classification+of+clinical+free+text%2E+In+) 13. Sampo Pyysalo, Filip Ginter, Juho Heimonen, Jari Björne, Jorma Boberg, Jouni Järvinen, and Tapio Salakoski. 2007. BioInfer: a corpus for information extraction in the biomedical domain. [↗](https://doi.org/10.1186/1471-2105-8-50) 14. BMC Bioinformatics 2007, 8:50 doi:10.1186/1471-2105-8-50. [↗](https://doi.org/10.1186/1471-2105-8-50)