with Abby Smith Ph.D. of NORC, University of Chicago
Wednesday, February 28th, 2024 7:30 – 8:30 pm at Terra Linda High School’s Innovation Hub
Description: Entity resolution, or deduplication, is sorting out duplicates in social networks. Picture cleaning up your friend list by comparing names and ages of potential duplicates (e.g. Abigail Smith vs. Abby Smith) . New methods group similar entities together, using friends’ connections to help. But here’s the thing: it’s not perfect, especially without a perfect list to compare to. Abby’s research looks into how good our sorting is, how to tweak the settings, and how these affect the bigger picture of the network. She tested this on real-world groups in Kentucky and Chicago, making sure sorting didn’t distort overall network inferences. In this presentation she will share how this work can be used for social good. Questions welcome!
Bio: Abby Smith recently received her Ph.D. in Statistics from Northwestern (2023) and is currently a statistician at NORC, working to improve entity resolution in big national opinion surveys. With a BS in Mathematics (2016) and an MS in Statistics (2017) from Carnegie Mellon, she’s driven by data science for social impact. As a 2022 DSSG fellow at CMU, Abby collaborated on a homelessness prevention project with Allegheny County DHS. Outside of stats, she enjoys running, backpacking, and has explored all 50 US states.(Fun fact! Abby and Alfia met through Postcrossing!)
Links: