Leveraging Temporal Trends for Training Contextual Word Embeddings to Address Bias in Biomedical Applications: Development Study.

ABSTRACT:

Background

Women have been underrepresented in clinical trials for many years. Machine-learning models trained on clinical trial abstracts may capture and amplify biases in the data. Specifically, word embeddings are models that enable representing words as vectors and are the building block of most natural language processing systems. If word embeddings are trained on clinical trial abstracts, predictive models that use the embeddings will exhibit gender performance gaps.

Objective

We aim to capture temporal trends in clinical trials through temporal distribution matching on contextual word embeddings (specifically, BERT) and explore its effect on the bias manifested in downstream tasks.

Methods

We present TeDi-BERT, a method to harness the temporal trend of increasi

SUBMITTER: Agmon S 

PROVIDER: S-EPMC11483253 | biostudies-literature | 2024 Oct

REPOSITORIES: biostudies-literature