Dhanya Sridhar: understanding what's human, one pattern at a time

Dhanya Sridhar, professor in Université de Montréal's Department of Computer Science and Operations Research

Dhanya Sridhar, professor in Université de Montréal's Department of Computer Science and Operations Research

Credit: Amélie Philibert | Université de Montréal

In 5 seconds

Through causality and machine learning, the Department of Computer Science and Operations Research professor is trying to better understand science and human behavior.

Does the style of language you hear help persuade you? Does texting with friends influence whether you vote? Does getting a politely written complaint make you respond faster.

These are the types of questions that Dhanya Sridhar, a new assistant professor in Université de Montréal's Department of Computer Science and Operations Research, is interested in.

A specialist in causality and machine learning, the researcher at Mila – Quebec Artificial Intelligence Institute has always had a keen interest in the patterns that underlie human behavior.

"I've always been very interested in understanding why complex behaviors and systems emerge as they do, and in learning about the reasons that motivate people to do what they do," she said.

Such scientific questions require reasoning about cause and effect, but determining causality is not always possible. In her work, Sridhar uses large-scale data and machine learning to uncover possible causal variables and uses the mathematical tools of causality to analyze these data.

In doing so, her goal is to understand, not just predict,  scientific phenomenon that govern our lives.

"The potential to answer and study more causal questions that affect people's lives and health through the use of machine learning and big data is what excites me," she said. "It's a long-term project, but it would be great to be able to demonstrate that these methods can provide solutions to pressing problems."

Her current research at UdeM and Mila focuses on fundamental research connecting causality and machine learning to build artificial intelligence that is more robust and can aid in discovering scientific knowledge.

During her postdoctoral and PhD research (at Columbia University and University of South California - Santa Cruz), she focused on computational social science applications, mainly in languages and conversations.

An American in Montreal

As someone who has always enjoyed languages, she is now able to apply this knowledge to her personal life.

Since moving from New York City to Montreal for her new job, Sridhar, an American citizen, has been learning French enjoying the opportunities she gets in everyday life in the city to speak and write in French.

Teaching students and graduate students, Sridhar tries to nurture independent thinking. Rather than repeating formulas that must be learned by rote, she prefers an approach that teaches her students  techniques they can then develop on their own.

"I want my students to feel creative and independent as much as they can," she said. "Research and graduate school are not always easy environments, but I think that as long as they can come away with a deep understanding of the field and an ability to make connections, it's possible for them to be better equipped to deal with future setbacks."

Always striving for a better understanding of what she teaches, the professor also tries to explain concepts as simply as possible, whether in her courses, lectures, or scientific publications.

"With experience, I have noticed that the more deeply we understand a subject, the more clearly we can explain it. So, when in doubt, it's important to ask ourselves why we are stuck and to keep going deeper and deeper."