How machine learning and data science facilitate COVID-19 research
Information has the potential to impact policy decisions, research directions, and personal decisions alike. Information-overload challenges informed and timely decision-making by increasing the chance of error and confounds decision-making. The key is to ensure relevant, actionable, and precise information is available concisely in near real-time.
To help, we formed a team of researchers and AI practitioners from Synaptiq, Karotene, and University of Maryland, Baltimore County (UMBC). The team is performing pro-bono research to address the need for clear understanding about the novel coronavirus Covid-19. The team will apply machine learning and data science to mine knowledge along these three research tracks:
- Published research.
- Social media content.
- Covid-19 statistics.
The first track, analyzing published research on the coronavirus pandemic, will cater to the knowledge needs of clinicians and scientists on the front-lines. The goal is to enable them to discover information from published research that is relevant to their work. This will be enabled by using automated, AI-driven knowledge discovery technology that has been under development at Karotene.
The second track, analyzing widely circulated content and social narratives about the ongoing pandemic on social media, will cater to the general public trying to stay informed and make sound decisions. The goal is to enable the general public to distinguish fact-based content from inaccurate and otherwise opinionated or editorialized commentary.
The third track will analyze statistics related to the trajectory of the epidemic across the globe with a view to inform policy decisions via published research. The objective is to draw insights related to the evolution of intervention strategies and public domain narratives and their impact on the pandemic’s trajectory.
The evolving Covid-19 pandemic has resulted in an explosion of information. Scientists, clinicians, policymakers, and the general public want accurate, concise, and actionable knowledge in real-time from ever-changing data. Scientists and clinicians are attempting to stay abreast and make sense of frantically reported global research data. This same content inundates the general public through news media, social media and personal networking channels.
Our combined team will be tackling this issue over the next several months in an effort to provide “social good” to our community. We will exemplify how machine learning and data science applied to public data can inform timely and critical decision-making in a time of strife. We will also share our Python notebooks and datasets online to encourage your collaboration
On April 7 at 12:30pm ET Dr. Tim Oates, Synaptiq’s co-founder and Chief Data Scientist, will present some of our early learnings through our BrightTalk Webinar channel at April 7. We hope you’ll join us.
By guest author Akshay Peshave