Trends in COVID-19 Publications: Streamlining Research Using NLP and LDA.

Akash Gupta,Anjali Agrawal,Himanshu Gupta,Shrey Aeron

doi:10.3389/fdgth.2021.686720

Abstract

Background: Research publications related to the novel coronavirus disease COVID-19 are rapidly increasing. However, current online literature hubs, even with artificial intelligence, are limited in identifying the complexity of COVID-19 research topics. We developed a comprehensive Latent Dirichlet Allocation (LDA) model with 25 topics using natural language processing (NLP) techniques on PubMed® research articles about “COVID.” We propose a novel methodology to develop and visualise temporal trends, and improve existing online literature hubs.Our results for temporal evolution demonstrate interesting trends, for example, the prominence of “Mental Health” and “Socioeconomic Impact” increased, “Genome Sequence” decreased, and “Epidemiology” remained relatively constant. Applying our methodology to LitCovid, a literature hub from the National Center for Biotechnology Information, we improved the breadth and depth of research topics by subdividing their pre-existing categories. Our topic model demonstrates that research on “masks” and “Personal Protective Equipment (PPE)” is skewed toward clinical applications with a lack of population-based epidemiological research.

Highlights

The COVID-19 outbreak was officially declared a pandemic by the World Health Organization in March 2020 [1]
To evaluate the temporal trends, we propose a novel method, which is applied to both PubMed R and LitCovid abstracts to produce an intuitive visualisation of the weekly temporal evolution of topic proportions
We provide a generalisable natural language processing (NLP) methodology to extract abstracts from PubMed R, create an optimised Latent Dirichlet Allocation (LDA) topic model, and visualise temporal trends

Summary

Introduction

The COVID-19 outbreak was officially declared a pandemic by the World Health Organization in March 2020 [1]. Latent Dirichlet Allocation (LDA) is an unsupervised topic modelling technique used to learn hidden topics within a corpus [2]. It assumes topics are a soft clustering of words and outputs two probability distributions: a distribution of topics in the corpus, and distributions of words across each topic. Current online literature hubs, even with artificial intelligence, are limited in identifying the complexity of COVID-19 research topics. We developed a comprehensive Latent Dirichlet Allocation (LDA) model with 25 topics using natural language processing (NLP) techniques on PubMed® research articles about “COVID.” We propose a novel methodology to develop and visualise temporal trends, and improve existing online literature hubs. Our topic model demonstrates that research on “masks” and “Personal Protective Equipment (PPE)” is skewed toward clinical applications with a lack of population-based epidemiological research

Methods

Results

Conclusion