The Arizona State University (ASU) Library’s Data Science and Analytics unit has developed a tool to accelerate COVID-19 research by helping researchers efficiently browse and understand the vast amount of research being published.
COVID-19 research is produced at such a rapid pace that it is challenging for researchers to keep up with the latest information, and to find new research results related to their specific focus. The Kaggle data-science community is collecting and distributing access to journal articles that report the results of COVID-19 research. The collection of articles is immense and expanding, so Kaggle challenged its collaborators to develop solutions that will help researchers stay up to date.
With guidance from Data Science and Analytics unit director Michael Simeone, research assistant and graduate student Steve Jadav developed a COVID-19 research-browsing tool to tackle this challenge within a few weeks. Parts of the application drew on previous work by research assistant Shashank Kapoor.
The COVID-19 research-browsing tool provides easy access to the curated set of journal articles and offers key analytic features for users to parse the collection. For example, users can conduct custom searches across all the articles as well as within individual articles. Given the massive size of the collection, the search algorithm is specifically programmed to return the most relevant results based on the user’s search terms.
The tool also uses natural language processing to generate on-demand article summaries. The ability to browse these summaries saves researchers time when they are conducting a preliminary analysis.
Additionally, users can download a list of relevant articles and their corresponding metadata, such as titles and authors, to formats suitable for custom analysis (CSV and Excel). Users can then conduct a Google search for a particular research article and additional information surrounding that article.
Since ASU launched the COVID-19 research-browsing tool at the end of April, more than 2,000 unique users have visited the site, retrieving over 350,000 articles in searches. The Data Science and Analytics unit is also working with researchers at the ASU Biodesign Institute to perform more specialized queries that can aid research on the half-life of the virus in various environments and situations. This work is a vital contribution to the global fight against COVID-19.