Skip to main content
U.S. flag

An official website of the United States government

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

Scientific Statement Classification over



Bruce R. Miller, Deyan Ginev


We introduce a new classification task for scientific statements and release a large-scale dataset for supervised learning. Our resource is derived from a machine- readable representation of the collection of preprint articles. We explore fifty author-annotated categories and empirically motivate a task design of grouping 10.5 million annotated paragraphs into thirteen classes. We demonstrate the task setup aligns with known success rates from the state of art, peaking at a 0.91 F1- score via a BiLSTM encoder-decoder model. Additionally, we introduce a lexeme serialization for mathematical formulas, and demonstrate that context-aware models could improve when also trained on the symbolic modality. Finally, we discuss the limitations of both data and task design, and outline potential directions towards increasingly complex models of scientific discourse, beyond isolated statements.
Proceedings Title
Language Resources and Evaluation Conference (LREC2020)
Conference Dates
May 3-5, 2020
Conference Location


natural language processing, deep learning, mathematical documents, arXiv


Miller, B. and Ginev, D. (2019), Scientific Statement Classification over, Language Resources and Evaluation Conference (LREC2020), Marseille, -1, [online], (Accessed June 18, 2024)


If you have any questions about this publication or are having problems accessing it, please contact

Created August 28, 2019, Updated September 25, 2020