RESEARCH

PROJECT STREAMS

Unmasking Sex Trafficking Supply Chains

In collaboration with the TellFinder Alliance, we have been exploring novel active learning approaches for uncovering trafficking risk in commercial sex supply chains using data from the deep web.  Mapping deceptive recruitment to sex-sales allows for new views into risky geographical corridors that can be used to inform interventions, coordination strategies, and investigations. Learn more here.

Analyzing Responsible Sourcing Strategies

We investigate the spillover effects from ethical certification programs (e.g. Fair Trade) into companies' non-certified supply chains. We connect 6 different structured and unstructured datasets for a unique view of coffee supply chains. We find that companies certifying as low as 3% of their portfolios experience positive spillovers in the rest of their supply chain.

Responsible AI

In collaboration with PricewaterhouseCoopers, we are designing a set of tools, games, and training content on key concepts related to responsible AI. We are working towards evaluating how these different tools impact ethical decision making through behavioral experiments with practitioners. Our goal is to find effective, practical methods for instilling ethical concepts in the daily processes of an organization.

WORKING PAPERS & PUBLICATIONS


Ramchandani, P., Bastani, H., and Wyatt, E.  Unmasking recruitment in sex trafficking supply chains with machine learning

Major Revision, M&SOM


Ramchandani, P., Bastani, H., and Moon, K. Responsible Sourcing: The First Step Is the Hardest

R&R Management Science


Ramchandani, P. and Bastani, H. “Multi-modal Active Learning for Network Discovery.”