Identifying COVID-19 cases and extracting patient reported symptoms from Reddit using natural language processing

Abstract We used social media data from “covid19positive” subreddit, from 03/2020 to 03/2022 to identify COVID-19 cases and extract their reported symptoms automatically using natural language processing (NLP). We trained a Bidirectional Encoder Representations from Transformers classification model...

Full description

Bibliographic Details
Main Authors: Muzhe Guo, Yong Ma, Efe Eworuke, Melissa Khashei, Jaejoon Song, Yueqin Zhao, Fang Jin
Format: Article
Language:English
Published: Nature Portfolio 2023-08-01
Series:Scientific Reports
Online Access:https://doi.org/10.1038/s41598-023-39986-7