The JimiLab Summer ’19 Plan

Many STEM faculty members at liberal arts college typically invite one or two undergraduate students to do summer research with them. This is not part of our teaching load but working with students on research projects can be exciting and productive.

For the past few years, I’ve typically had a big summer crew of 4-6 students work with me on my project. We did the “tech startup in college setting” thing in which I acted a bit like a CEO and the students were UI/UX designers, software engineers, and marketing coordinators. However, during my 2017-18 sabbatical year in Singapore, I decided to shut MegsRadio down after 8 years and two NSF grants. And having been abroad for the year, my undergraduate research lab was decimated so I did not have any summer research students last year for the first time in my 10 years as being a professor. It was a productive summer of solo work but it was also kinda lonely from a research perspective.

So this summer, I over compensated. I had initially planned to work with 2 students on one research project involve local music recommendation. However, as the spring semester progressed and more funding became available, the number of summer students in my lab kept creeping up. Currently, I have ten students working with me on six different projects. I’m pretty excited about it right now but fear I could get overwhelmed if I am not careful about making sure that each student learns to be self-motivated and self-sufficient.

Here is what is going on:

Localify (with Tim Clerico, Erich Ostendarp, Nic Wand, Vianca Hurtado, collaboration with Thorsten Joachims): an app development project that looks at how we can use Spotify to promote music and events from local artists.

Local (Long-Tail) Music Recommendation (with Gabe Pesco, collaboration with Danny Akimchuk, Ted Galanthay, Thorsten Joachims): Recommender systems typically recommend popular music due to both commercial and technical biases. How can we reverse these biases and recommend music by talented, less well-known, local artists?

Venue-Artist Event Selection (with Josh Hayden, collaboration with Doug Levine): We are working with the State Theater of Ithaca (and other historical theaters) to help increase participation in events by helping to find exciting artists.  Can we use data to help venues find artists and to help artist find venues?

Porchfest App (with Yemi Afolabi, collaboration with Alex Python): Porchfest is a simple idea: provide a time and a place where local musicians can play music on their porch. Listeners can wander the street and discovery amazingly talented local musicians. We are building an app to support this kind of locally-focused music discovery. 

Synthesizer Reverse Engineering (with Joe Cleveland, Arshiya Gupta): We hear many different synthesized instruments in music today. If we want to perform these songs, we need to figure out which audio filters are used and how to set the parameters for them.

Broadway Ticketing Analytics (with Alex Python, collaboration with Justin Karr): consulting with Jujamcyn Theaters to help with various data analytics problem related to Broadway ticket sales.

The first four project are related to my interested in helping to promote music by local artists. This continues to interest me despite the fact that MegsRadio never got off the ground. One way that I am trying to learn from that experience to focus a bit more on initial research (as opposed to full-blown development) and have multiple student-driven projects so that we do not have all of our eggs in one basket.

It will be a power-packed summer for sure. Stay tuned.