Purohit Working On Project Examining Use Of Big Data Analytics To Aid COVID-19 Response
Hemant Purohit, Assistant Professor, Information Sciences and Technology, received $24,266 from the National Science Foundation for a project in which he and his collaborators are implementing a framework for social media use in disasters in a computational system.
The researchers developed this framework as part of a previous research project, and will further implement the framework for this effort. They are taking this step to discern how content affects the trust of human analysts who are teaming with artificial intelligence (AI) systems and how to design more efficient ways to improve the collaboration between humans and AI systems for complex information processing tasks.
Purohit will also provide the infrastructure of the stream processing system CitizenHelper resulting from the previous project to support the social media stream collection, storage, and real-time analysis using machine-learning algorithms with an active learning interface.
Further, he will conduct empirical analyses of different sampling algorithms for active learning methods to develop a topology of errors across different contexts of the sampled content and determine how the contexts affect the quality of human decision-making for feedback.
He will also engage in outreach with an existing AI and disaster management practitioner community.
Funding for this work began in May 2020 and will end in late April 2021.
###