Active Learning Yields Better Training Data for Scientific Named Entity Recognition

Published

November 1, 2019

Author(s)

Roselyne B. Tchoua, Aswathy Ajith, Zhi Hong, Logan T. Ward, Kyle Chard, Debra Audus, Shrayesh N. Patel, Juan J. de Pablo

Abstract

Despite significant progress in natural language processing, machine learning models require substantial expert-annotated training data to perform well in tasks such as named entity recognition (NER) and entity relations extraction. Furthermore, NER is often more complicated when working with scientific text. For example, in polymer science, chemical structure may be encoded using nonstandard naming conventions, the same concept can be expressed using many different terms (synonymy), and authors may refer to polymers with ad-hoc labels. These challenges, which are not unique to polymer science, make it difficult to generate training data, as specialized skills are needed to label text correctly. We have previously designed polyNER, a semi-automated system for efficient identification of scientific entities in text. PolyNER applies word embedding models to generate entity-rich corpora for productive expert labeling, and then uses the resulting labeled data to bootstrap a context-based classifier. PolyNER facilitates a labeling process that is otherwise tedious and expensive. Here, we use active learning to efficiently obtain more annotations from experts and improve performance. Our approach requires just five hours of expert time to achieve discrimination capacity comparable to that of a state-of-the-art chemical NER toolkit.

Proceedings Title

2019 IEEE 15th International Conference on e-Science (e-Science)

Conference Dates

September 24-27, 2019

Conference Location

San Diego, CA, US

Pub Type

Conferences

Download Paper

DOI Link

Citation

Tchoua, R. , Ajith, A. , Hong, Z. , Ward, L. , Chard, K. , Audus, D. , Patel, S. and de Pablo, J. (2019), Active Learning Yields Better Training Data for Scientific Named Entity Recognition, 2019 IEEE 15th International Conference on e-Science (e-Science), San Diego, CA, US, [online], https://doi.org/10.1109/eScience.2019.00-71 (Accessed October 12, 2025)