GDIT Joins Tech Community and White House to Mine Scientific Research for a Global Challenge
As the world continues to confront the COVID-19 pandemic, we find ourselves collectively in an environment that demands, more than ever, quick action, expertise and collaboration. GDIT has joined a White House Artificial Intelligence Challenge and the scientific community to find answers to big questions on COVID-19.
Why does the virus affect different people in dramatically different ways?
Enter the COVID-19 Open Research Dataset (CORD-19). The White House and a coalition of research groups including the Allen Institute and Google (with its data science subsidiary Kaggle), collected more than 135,000 machine-readable scholarly articles about coronaviruses and made it accessible to the entire global research community.
The challenge calls for the nation’s AI experts to develop new text and data mining techniques using the open dataset. By applying scientific understanding combined with new advances in natural language processing (NLP), machine learning (ML) and AI, GDIT is responding to the challenge to help find new insights about the virus and support the worldwide effort to eradicate it.
Natural Language Processing Analysis Links Diabetes Risks
A GDIT team was assembled to focus on finding answers to COVID-19 risk factors, including in-house epidemiologists, NLP and AI/ML experts and those supporting missions with populations with chronic conditions, who may be among those most impacted by the virus. By examining the prevalence and co-occurrence of number of patients, comorbidities and percentages, the team started to find linkages between diabetes and COVID-19 deaths and the specific risks the virus presents for people with diabetes.
This initial effort, and the team’s method of extracting and understanding the entities mentioned, has provided a framework to not only look at other underlying health conditions (such as hypertension, heart disease, etc.) but to also expand the scope of entities examined to include things like geographic or demographic data as well – e.g. how old are the patients, what’s their ethnicity, where do they live and why could that be important?
The team’s participation in the challenge was made possible by the fact that GDIT had in-house experts in NLP and AI/ML, knowledge of a specific population of COVID-19 patients and the support internally from company leadership to focus on this effort. Its work will soon be published in scientific journals and will continue to expand, building on its initial submission.
In the coming months, the GDIT team will continue to work together in support of this critical global health sciences need.