Can We Compute Society?

About the Lecture

Can a computer detect political polarization? Can an algorithm tell whether a news outlet is biased?

Can we map an online community, track the spread of misinformation, or teach a machine to understand what people are arguing about? Computational Social Science sits at the curious intersection of people and machines, using data, algorithms and AI to study everything from elections and social media to culture, language and collective behaviour.

In this lecture, Professor Anirban Sen takes us on a tour of the field's most fascinating — and occasionally surprising — applications, drawing on his own research into misinformation, hate speech and propaganda communities online. Along the way, we'll ask the question underneath all of it: what do we gain, and what do we lose, when we turn society into data?

The session explores:

How researchers use data and algorithms to detect polarization and media bias
What happens to misinformation, hate speech and propaganda communities as they grow and change online
The techniques behind the field: NLP, machine learning, information retrieval and good old qualitative analysis
What it means to study people at the scale of the entire web, and where that approach runs out of road
About the Speaker
Anirban Sen is an Assistant Professor in the Department of Computer Science at Ashoka University. Before joining Ashoka in 2024, he was a postdoctoral researcher at Microsoft Research India, on the Technology for Emerging Markets team, where he studied misinformation, hate speech and propaganda communities on social media and how they evolve over time.

He completed his PhD in Computer Science and Engineering at IIT Delhi, where he was part of the Appropriate Computing Technologies for Development (ACT4D) group and helped build a system for political economy analysis of Indian policy events using publicly available web and media data. He has also worked in industry as a data scientist, building solutions across banking, finance and insurance, including with generative AI.