Keyword extraction in BERT-based models for reviewer system
Josh Newburn,
Loading…
Permalink
https://hdl.handle.net/2142/121148
Description
Title
Keyword extraction in BERT-based models for reviewer system
Author(s)
Josh Newburn,
Issue Date
2023-05-01
Keyword(s)
Keyword extraction, BERT, deep learning, natural language processing
Abstract
For keyword extraction tasks to obtain high quality phrases that encapsulate the overall meaning of a sentence, paragraph, or document, shallow systems have been employed to obtain satisfactory results. In this project, we research and implement methods to use deep learning methods to improve upon
these systems for keyword extraction. We find that a pretrained BERT-model passed through the KeyBERT keyword extractor is effective at obtaining keywords from any given document, and we improve upon this result by fine-tuning a BERT-based model using token classification from pretrained bert-base-uncased and Adam optimization. We generate a larger quantity of high-quality phrases than created by the shallow system algorithm, and these keywords give better insight into the context of a paper within a specific domain gave by providing more descriptive keywords.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
