JMIR Publications

ABSTRACT

Background:

Evidence from peer-reviewed literature is the cornerstone for designing responses to global threats such as COVID-19. The collection of knowledge in publications needs to be distilled into evidence by leveraging natural language models and machine learning.

Objective:

We aim to show that new knowledge can be captured and tracked using the temporal change in the underlying unsupervised word embeddings of literature. Further imminent themes can be predicted using machine learning upon the evolving associations between words.

Methods:

Frequently occurring medical entities were extracted from the abstracts of more than 150,000 COVID-19 articles published on the WHO database, collected on a monthly interval starting from February 2020. Word embeddings trained on each month's literature were used to construct networks of entities with cosine similarities as edge weights. Topological features of the subsequent month’s network were forecasted based on prior patterns and new links were predicted using supervised machine learning. Community detection and alluvial diagrams were used to track biomedical themes that evolved over the months.

Results:

We found that thromboembolic complications were detected as an emerging theme as early as August 2020. A shift towards symptoms of Long COVID complications was observed during March 2021 and neurological complications gained significance in June 2021. A prospective validation of the link prediction models achieved an AUROC score of 0.87. Predictive modelling revealed predisposing conditions, symptoms, cross-infection and neurological complications as a dominant research theme in COVID-19 publications based on patterns observed in previous months.

Conclusions:

Machine learning-based prediction of emerging links can contribute towards steering research by capturing themes represented by groups of medical entities, based on patterns of semantic relationships over time.