Data visualization of Lorenz attractor system, also known as butterfly effect

Chakraborty awarded NSF grant to develop AI approaches for identifying dynamic clinical data patterns, causal links in text

October 2, 2020

Sunandan Chakraborty

Sunandan Chakraborty, assistant professor of data science at the IU School of Informatics and Computing at IUPUI, recently received a $174, 332 CRII award from the National Science Foundation’s Division of Information and Intelligent Systems. The grant project, entitled “Capturing Dynamism in Causal Relationships: A New Paradigm for Relationship Extraction from Text” will use artificial intelligence (AI) to discover new clinical data patterns in vast repositories of text data.

Text mining is extensively used in medicine, health, economics, public policy, and journalism. This project aims to address public health and how changing climatic, political, and economic policies and conditions affect mental and physical health of populations in different geographic areas.

The AI technique has been used to convert large, unstructured text data into knowledge, but has been limited to capturing snapshots in time, rather than the dynamic nature inherent in knowledge. Findings may be conflicting, inconsistent, refuted, contradicted, or confirmed over time. This project intends, in particular, to look at any causal relationships, i.e., changes in one variable resulting from changes in another variable.

Chakraborty’s research will be used to discover little known symptoms and signals of the rare disease Sjögren’s Syndrome by mining research articles and clinical notes in order to target significant reduction of diagnosis time for this disease. This process can be used to study other diseases as well.

He is also collaborating with the CDC to “identify the latent connections between real-world events and mental health issues of the general population by mining and validating causal statements from scholarly medical text, news, and social media data.” The CDC investigators are particularly interested in detecting the causal effects of extreme weather conditions, e.g. heat wave, or drastic changes in the economy, e.g. sudden loss of jobs, on mental health issues.

Although Chakraborty and his colleagues originated this research prior to the current COVID-19 pandemic, they are now thinking about how this project can help in this and similar public health crises.

Davide Bolchini, chair of the Department of Human-Centered Computing and professor of human-centered computing at the School of Informatics and Computing, said, “Sunandan’s project is very timely and original, because it augments text mining with state-of-the art neural network approaches in order to drive more powerful tools that can extract hidden meaning from text data, especially clinical information. This is an exciting example of research that is advancing the fundamentals of computing while creating the basis to make a positive impact on public health.”


This material is based upon work supported by the National Science Foundation under Grant No. 1948322. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.​

Media Contact

Joanne Lovrinic