Human-Computer Interaction Institute
Carnegie Mellon University
5000 Forbes Avenue
Pittsburgh, PA 15213

Office: 4604 Newell Simon Hall
Twitter: @mmadaio

Who I Am

I am currently a PhD student in the Human-Computer Interaction Institute at Carnegie Mellon University. I am advised by Drs. Justine Cassell and Amy Ogan, and I currently work in the ArticuLab, studying the impact of social relationships on learning.

Before coming to Carnegie Mellon, I completed a Masters of Science in Digital Media at Georgia Tech, advised by Dr. Ian Bogost.

Prior to that, I graduated from the University of Maryland with a Masters of Education and a Bachelors of Arts in Language and Literature, and taught for several years at a public high school in Maryland.

In my free time (whatever that is), I enjoy writing and teaching music, writing and performing slam poetry, and exploring the outdoors.

Research Interests

I am primarily interested in the ways that technology mediates social learning interactions, between students and their peers, their instructor, or a virtual tutoring agent.

Currently, I study how rapport between peer tutors impacts the learning process and outcomes, and we use those findings to develop a virtual tutoring agent that can build rapport with the students who use it, to better motivate their learning.

Along related lines, I am interested in how computational tools such as machine learning and AI can augment civic decision-making processes. To that end, I am currently a project lead with Metro21, a public-private partnership between CMU and the City of Pittsburgh's Department of Innovation and Performance.

Current Research

Rapport-Aligned Peer Tutor

Interpersonal rapport is an important element of human interaction, particularly in educational contexts, but current educational technologies often ignore the importance of interpersonal social interaction in learning. In this work, we are developing a computational model of rapport, to incorporate into an intelligent tutoring system, to both better motivate the students that work with it, and to empirically validate theories of rapport.

A paper describing early findings from this work, on the different tutoring and learning styles between friends and strangers, and their impact on learning, has been published in the 2016 Intelligent Tutoring System conference.

Previous Projects

Firebird: Predicting Fire Risk

The Firebird framework was designed to augment municipal fire departments' commercial property fire inspections through a predictive model of fire risk for commercial properties and the visualization of historical fires and property inspections on an interactive map. This project was a collaboration between Georgia Tech and the Atlanta Fire Rescue Department through the Data Science for Social Good program.

Read more about this work in a paper accepted to the Knowledge Discovery and Data Mining (KDD) 2016 conference, and a shorter paper presented at the Bloomberg Data for Good Exchange in 2015.

iLab Liberia

To better understand whether the educational technology I was researching at Georgia Tech would be beneficial in introducing a global audience to computing, I spent the summer of 2014 teaching introductory computer science classes at the iLab Liberia.

Read a paper from this research on developing sustainable communities of practice for computer science education in low-resource areas, and a paper from another strand of this research on social motivations for global students taking online classes, or MOOCs, both published in the ICT for Development (ICTD) conference, in 2015 and 2016, respectively.

EarSketch Curriculum

EarSketch is a browser-based programming platform that builds on the commonalities between computer science and music. By writing a script in Python or Javascript, students are also writing songs, using musical samples as variables and constants, looping and modifying them through the functions they write.

On this project, I worked with the curriculum development team, helping music teachers understand enough programming to use this in their classes, and helping computer science teachers understand enough music to use this to motivate students.

EarSketch Android App

This project involved the design and development of an Android app for EarSketch, a platform that uses digital music composition to teach programming concepts and computational thinking skills.

This app supplements the main EarSketch IDE by enabling collaborative peer programming in informal learning settings. It uses a block-based metaphor to allow beginner programming students to understand concepts like variables, functions, recursion, and iteration.

Streetlight Illuminance Mapping

As a different avenue of data analytics for social good, this project measured the brightness of walking paths on Georgia Tech's campus, and mapped those brightness values onto GT police crime report data to reveal connections between crime incidents and poorly lit areas on campus.

If you want to make your own light map of your area, check out our DIY instructions on Instructables.


This was a class project in the Digital Media program at Georgia Tech, a mockup of an app for teachers to dynamically grade student participation in live class discussions.

Teachers would be able to use this app to give credit for student participation quickly and easily, while also moderating a live discussion.