Identifying fraud victims to account for, and eliminate bias in machine learning.
What makes some groups of people more vulnerable to online fraud? How do fraud prevention measures reflect the psychology of the victim? These are some of the questions we tackle in the newest episode of The Full Cart, Riskified’s podcast about the eCommerce landscape.
This week, we sat down with Reka Eszter Bodo, a research analyst on Riskified’s model training team. Reka says that protecting marginalized communities from social engineering and fraud is a matter of moral responsibility, not just a business’ bottom line. Finally, Reka illustrates how she uses her insights on fraud to train better, less biased ML fraud detection models.
Tune in for insights on:
- Which demographics are most vulnerable to online fraud – and why that matters to merchants
- How customer service teams can play a key role in making online shopping safer
- Why human insights are crucial to creating better machine learning models
If you missed our last episode, From Crisis to Transformation to Stabilization: “What’s happening in eCommerce throughout the Coronavirus Crisis?” you can listen to it here.