Transfer Learning from BERT to Support Insertion of New Concepts into SNOMED CT
Document Type
Article
Publication Date
1-1-2019
Abstract
With advances in Machine Learning (ML), neural network-based methods, such as Convolutional/Recurrent Neural Networks, have been proposed to assist terminology curators in the development and maintenance of terminologies. Bidirectional Encoder Representations from Transformers (BERT), a new language representation model, obtains state-of-the-art results on a wide array of general English NLP tasks. We explore BERT's applicability to medical terminology-related tasks. Utilizing the "next sentence prediction" capability of BERT, we show that the Fine-tuning strategy of Transfer Learning (TL) from the BERTBASE model can address a challenging problem in automatic terminology enrichment - insertion of new concepts. Adding a pre-training strategy enhances the results. We apply our strategies to the two largest hierarchies of SNOMED CT, with one release as training data and the following release as test data. The performance of the combined two proposed TL models achieves an average F1 score of 0.85 and 0.86 for the two hierarchies, respectively.
Identifier
85083755680 (Scopus)
Publication Title
AMIA Annual Symposium Proceedings AMIA Symposium
e-ISSN
1942597X
PubMed ID
32308910
First Page
1129
Last Page
1138
Volume
2019
Recommended Citation
Liu, Hao; Perl, Yehoshua; and Geller, James, "Transfer Learning from BERT to Support Insertion of New Concepts into SNOMED CT" (2019). Faculty Publications. 7928.
https://digitalcommons.njit.edu/fac_pubs/7928
