Patricia Van Zandt
$299,836
Sujan Ranjan Re Anreddy
Terri N Hernandez
Mississippi State University
Mississippi
Social Behavioral and Economic Sciences (SBE)
Human Networks & Data Sci Infr
The COVID-19 Online Prevalence of Emotions in Institutions (COPE-ID) database contains online discussions of COVID-19, including posts about emotions, such as fear, anxiety, and social institutions, such as healthcare and family. These data can be used to answer questions about the spread of information and individual well-being during the COVID-19 pandemic. This project creates a data visualization tool to process social media data from COPE-ID. This tool makes it easier for people to explore large volumes of social media data to study the emotions, thoughts, behaviors, and health of people during a pandemic or related disaster. The visualization tool allows researchers of all backgrounds and skill levels to access and process data from the COPE-ID, as well as data from other social media sources. Improving access to COPE-ID data can inform future public health policies and interventions. The data visualization tool’s users will be able to access an overview of large social media datasets through a platform dashboard. The dashboard presents visualizations of the data that are constructed by topic modeling algorithms, which produce a summary of the data in the form of word and topic frequencies. The tool also allows users to perform sentiment analysis, such as the attitude toward topics from negative to positive. Visualizations such as word clouds and time series charts generate insights for users to drive their task-based interactions with the tool. Users can also request samples of data that can be labeled using qualitative or content analysis. This labeled data can then be used to make predictions about future events, predictions that are generated by advanced statistical analyses or machine learning techniques. Training datasets can be used to code and process COPE-ID data, and these coded datasets can be used to examine the rate of agreement between coders so that the quality of the data can be improved. The tool improves scientists' access to social media data and allow researchers to test theories of human behavior using user generated big data. This project is jointly funded by Human Networks and Data Science -- Infrastructure (HNDS-I) and the Established Program to Stimulate Competitive Research (EPSCoR). This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.