TY - GEN
T1 - Writing strategies for science communication
T2 - 2020 Conference on Empirical Methods in Natural Language Processing, EMNLP 2020
AU - August, Tal
AU - Kim, Lauren
AU - Reinecke, Katharina
AU - Smith, Noah A.
N1 - We thank Kristin Osborne, Rebekka Coakley, and Sarah E. McQuate for their helpful input on science communication, the annotators for their work on the project and the anonymous reviewers and members of the UWNLP community for their helpful feedback. This work was supported in part by the Office of Naval Research under MURI grant N00014-18-1-2670.
PY - 2020
Y1 - 2020
N2 - Communicating complex scientific ideas without misleading or overwhelming the public is challenging. While science communication guides exist, they rarely offer empirical evidence for how their strategies are used in practice. Writing strategies that can be automatically recognized could greatly support science communication efforts by enabling tools to detect and suggest strategies for writers. We compile a set of writing strategies drawn from a wide range of prescriptive sources and develop an annotation scheme allowing humans to recognize them. We collect a corpus of 128K science writing documents in English and annotate a subset of this corpus. We use the annotations to train transformer-based classifiers and measure the strategies' use in the larger corpus. We find that the use of strategies, such as storytelling and emphasizing the most important findings, varies significantly across publications with different reader audiences.
AB - Communicating complex scientific ideas without misleading or overwhelming the public is challenging. While science communication guides exist, they rarely offer empirical evidence for how their strategies are used in practice. Writing strategies that can be automatically recognized could greatly support science communication efforts by enabling tools to detect and suggest strategies for writers. We compile a set of writing strategies drawn from a wide range of prescriptive sources and develop an annotation scheme allowing humans to recognize them. We collect a corpus of 128K science writing documents in English and annotate a subset of this corpus. We use the annotations to train transformer-based classifiers and measure the strategies' use in the larger corpus. We find that the use of strategies, such as storytelling and emphasizing the most important findings, varies significantly across publications with different reader audiences.
UR - https://www.scopus.com/pages/publications/85101799785
UR - https://www.scopus.com/pages/publications/85101799785#tab=citedBy
U2 - 10.18653/v1/2020.emnlp-main.429
DO - 10.18653/v1/2020.emnlp-main.429
M3 - Conference contribution
AN - SCOPUS:85101799785
T3 - EMNLP 2020 - 2020 Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference
SP - 5327
EP - 5344
BT - EMNLP 2020 - 2020 Conference on Empirical Methods in Natural Language Processing, Proceedings of the Conference
PB - Association for Computational Linguistics (ACL)
Y2 - 16 November 2020 through 20 November 2020
ER -