I am a fifth-year Ph.D. student in the Machine Learning Department at Carnegie Mellon University, jointly advised by Jen Mankoff, in the Human-Computer Interaction Institute, and Steve Fienberg, in the Department of Statistics. My research interests are, broadly, using computation to solve challenges in environmental sustainability and intelligently ordering questions in online surveys. I am a member of the Make Abilities Group.

Outside of research, I’m also involved with Women@SCS at CMU and volunteer with TechNights, a weekly program to introduce middle-school girls to the excitement of computer science through hands-on lessons and activities. Examples of sessions I helped to design and lead include recommender systems, parallelism, and signal processing.

Before coming to CMU in 2012, I got my undergraduate degree from Vanderbilt University in Nashville, Tennessee, where I majored in computer science, classics, and math.

Current Research Projects

EDigs logo

Easy, economical, efficient rental searches

EDigs is a system to help prospective tenants make decisions in their apartment searches by giving them quality-of-life information that often is not revealed until after they sign a lease and move in. With EDigs, users can rate apartments and landlords, take pictures and notes associated with apartments they have visited, and share these notes with roommates. My focus for EDigs has been on personalized energy estimates -- providing predictions for electricity and natural gas usage, based on physical features of a home and the behavior of the people who will live in it. If you're looking to rent in Pittsburgh, check out EDigs!

This is joint work with a number of other students, led by Jen Mankoff.

DQO illustration

Dynamic Question Ordering

Online surveys offer the possibility for adaptive questions, where later questions depend on responses to earlier questions. We present a general framework for dynamically ordering questions to engage respondents, improving survey completion and imputation of unknown items. Another problem for dynamic question ordering is giving personalized predictions. Since it is possible to give a reasonable prediction with only a subset of questions, we are not concerned with motivating the user to answer all questions. Instead, we want to order questions so that the user provides information that most reduces prediction uncertainty, while not being too burdensome to answer.

This is joint work with Jen Mankoff and Steve Fienberg.

Hark! a Vagrant cartoon of Ada Lovelace

Gender and Authorship

We are examining differences in gender as they relate to productivity, influence, and collaboration in HCI publications over time. From bibliometric information, we analyze the career trajectories of female and male researchers in computer science. Our approach compares people to their past selves, which eliminates individual- and personality-related factors that influence success as well as the selection bias for individuals at higher ranks. Because our metrics can be extracted from publication records alone, this analysis can be easily repeated for different groups of authors and for different timelines to see the impact of gender on different fields at different times.

This is joint work with Jess Hammer, Jen Mankoff, and Anna Wong.

Fun Facts

ring-juggling picture

I learned to juggle in undergrad.

Finding Nemo

Finding Nemo is the best movie.

me in the woods

The woods adore me.