In a groundbreaking development, a recent publication in Health Data Science, a partner journal of Science, introduces a sophisticated analytical framework. This framework, designed to navigate the extensive textual realm associated with COVID-19, incorporates keywords derived from Google Trends and abstracts from the WHO COVID-19 database. It offers a unique and detailed comprehension of the dynamic discourse surrounding the pandemic.
Throughout the crisis, research has been pivotal in shaping effective policy. However, tools like Google Trends frequently need to pay more attention to the intricate details that scholarly research captures. By comparing Google Trends data with academic articles, the study sheds light on the scope and intensity of scientific discourse on COVID-19 topics about public interest.
Benson Shu Yan Lam, an Associate Professor at The Hang Seng University of Hong Kong, emphasizes the study’s aim to examine the promptness and interconnectedness of these information sources, determining whether Google Trends can effectively indicate emerging public concerns or if scholarly discussions offer more immediate and thorough insights.
Amanda Man Ying Chu, an Assistant Professor at The Education University of Hong Kong, made a notable discovery: the academic community has precedence in tackling COVID-19 issues. Academic abstracts initiated discussions on these topics before they became prominent in Google Trends searches, providing deeper insights invaluable for policy development.
The research introduces the Coherent Topic Clustering (CTC) technique, a novel text-mining method that efficiently categorizes significant phrases from extensive research abstracts. This technique surpasses BERTopic, a modern deep-learning model, in identifying relevant themes.
Looking ahead, Mike Ka Pui So, a Professor at The Hong Kong University of Science and Technology, envisions a promising future for this analytical framework. He anticipates extending its application beyond health science to include financial news analysis. This potential expansion underscores the framework’s ability to integrate qualitative and quantitative insights, marking a significant advancement in the field of financial analysis.
Reflecting on the study’s broader implications, Professor So discusses the potential for combining various data types and integrating textual analysis with traditional numerical data sets to enhance financial analytics. This approach illustrates the research’s interdisciplinary possibilities, offering new pathways for comprehensive analysis across different fields.
The study’s findings reveal that research abstracts addressed the majority of COVID-19 topics before Google Trends highlighted them and offered a more comprehensive examination of these issues. That could significantly aid policymakers in recognizing the central issues related to COVID-19 and enable them to respond more promptly. Furthermore, the clustering technique more accurately captures the principal themes of the abstracts compared to a recent sophisticated deep learning-based approach to topic modelling. The academic community engages with COVID-19 topics more rapidly than Google Trends.
More information: Benson Shu Yan Lam et al, Do Scholars Respond Faster Than Google Trends in Discussing COVID-19 Issues? An Approach to Textual Big Data, Health Data Science. DOI: 10.34133/hds.0116
Journal information: Health Data Science
