|
Optimal’s education research and data science teams were represented at the 2022 International Population Data Linkage Network (IPDLN) conference at the University of Edinburgh in Scotland. Optimal's data science director presented on the use of data science methods and tools used to pilot real-time monitoring of school district COVID data.
To collect the data, research scientists, led by Sadaf Asrar, developed customized robotic process automation programs that used a combination of web scraping and Natural Language Processing (NLP) algorithms to collect data on COVID cases in schools, mask mandates at the school districts level, as well as information on school closures and mode of instruction. For instance, during the pilot phase of this effort, Optimal’s data scientists demonstrated that the trained algorithm could independently find and classify school districts that had indeed implemented a mask mandate at an accuracy of 87%. These early results demonstrate that selected school district-level core components can be monitored in real-time, at scale, and at lower costs. Such scans can be repeated frequently and can complement or help validate official statistics.
Optimal's Data Science and Education Projects
Optimal has worked with the U.S. Department of Education since 2003. We have collaborated with the National Center for Education Statistics to contextualize ED data by developing products such as the State Assessment and Education Policy (SAEP) tool. We have linked data on natural disasters, school safety, teacher compensation, along with COVID-19 information in the United States from the web by using tools powered by artificial intelligence to help contextualize the 2022 National Assessment of Educational Progress (NAEP) assessment results.
| 
