May 11, 2020
The Collaborative Archive & Data Research Environment (CADRE) is welcoming four fellow teams to its Research Cohort for the Study of Coronaviruses (RCSC).
The RCSC Program was created in response to the White House's call to action to help answer COVID-19 questions with the COVID-19 Open Research Dataset (CORD-19) of scholarly literature. CADRE’s RCSC fellows, who will be developing research related to COVID-19, will have access to the CORD-19 dataset on the platform, as well as to the Web of Science and Microsoft Academic Graph datasets.
Read more about the four new RCSC research teams below. While CADRE wants to support as many RCSC research projects as possible, the tean has put a pause on accepting new fellows until it has the capacity to provide intensive support for new teams. You can learn more about the RCSC Program here.
These researchers plan to address the recent flood of COVID-19 studies into preprint repositories. The sudden influx makes it difficult for researchers to stay on top of research that is relevant to their particular field using metrics or citations. The research team proposes a network-driven study that summarizes the fields related to recent COVID-19 literature. They will accomplish this by building citation networks among the most recent studies and applying community detection and keyword extraction to understand how the literature is organized. The researchers say that automated methods for summarizing COVID-19 research, as well as interactive visualizations, will aid other researchers in finding the most relevant and recent research in their field. The team will use CADRE's unified environment for access to scholarly datasets to accelerate their research.
This project will create a map of papers on COVID-19 that will be compared with maps for similar viruses, such as SARS and influenza, to better understand unexplored and concentration areas in the research. Kojaku will employ neural embedding methods, used to project research papers onto a low dimensional space using citations and semantic information, and then compare the density of papers on COVID-19 with other diseases. Kojaku says this research will both address unexplored areas that may deserve more attention and allow researchers to easily explore research related to COVID-19. Kojaku is part of the Science Genome project, which is a close CADRE collaborator, and says he will use CADRE to access the comprehensive, continuously updated CORD-19 dataset and related research, as well as to receive support from CADRE’s technical team and access to its computing resources.
This research team plans to study the international collaborations that are forming to perform COVID-19 research. The researchers say the combination of an intense research focus and a demand for quick results provides a rare opportunity for social scientists who study collaboration, teaming, and science dynamics. The team added that the abundant informal communications and knowledge sharing among scientists to address COVID-19 is rare. All of these circumstances have created an opportunity for the researchers to study team formations in real-time with CADRE. They plan to take advantage of CADRE’s datasets and technical team for their research. Two team members (Wagner and Cai) are also currently using CADRE for another CADRE Fellowship project.
This project will address the quality of recently published COVID-19 publications. Sevryugina says COVID-19 related research is being performed and published hastily. She adds that turnaround times for medical journal publications have decreased by almost 50 percent. Speedy research and condensed publication timelines contribute to a lack of scientific rigor and increase the likelihood of corrections and retractions, leading to the spread of false information in trusted journals. The team will study the quality of COVID-19 related scholarly works by using CADRE’s datasets to identify signs of incoherency, irreproducibility, and haste. That includes analyzing COVID-19 published literature for incoherent writing, stylistic errors, plagiarism, speculative language, unreproducible experiments, and far-fetched conclusions based on poor quality data. Researchers will also examine retracted and corrected manuscripts and explore their citation maps to understand how errors propagate through scholarly literature. The fellows hope their research will help others find the balance between expediting publication timelines and maintaining research quality.