SuperUROP All SuperUROP Projects (hass)
spacing
SuperUROP
 
Home
Letters of Recommendation
Login

SuperUROP Projects

"Live music listings app, web dev, marketing"
Faculty Advisor: Ian Condry
Mentor(s):
Contact e-mail: condry@mit.edu
Research Area(s):
We are a small team developing an app and website for live music listings in the Boston area. We have an alpha version of the app and the website which we are preparing to release by mid-summer. We are seeking students to work on the app, website, and marketing. This will involve working on a team, preferably 20-40 hours per week, but we are flexible. In addition to professor in Comparative Media Studies, team includes data entrepreneur, programmers, and tech consultant. Interest in live music a plus. Our goal is to enliven the local live music scene by helping people discover sounds they didn’t know they'd like.

Prerequisites: We are seeking people with app development and web development experience, but we are open to considering people with some programming experience and a desire to learn. We prefer those with experience with React, React Native, Java Script / Node JS, HTML, CSS, MySQL, and/or experience with apps or web dev for IOS and Android. We also seek people interested in working with local musicians and venues to improve outreach, user experience, social media engagement, and in-person networking.
"Generative Music with Machine Learning"
Faculty Advisor: Eran Egozy
Mentor(s):
Contact e-mail: egozy@mit.edu
Research Area(s):
This research project will explore the intersection of music and machine learning, in particular, by using ML algorithms to automatically generate musical compositions based on specific styles via learned examples.

We will look at which datasets can be used and what representational structures are most suitable for this task.

We will also attempt to go beyond "just generating notes", to addressing more musical parameters such as instrumentation and dynamics.
"Digital Humanities"
Faculty Advisor: Michael Scott Cuthbert
Mentor(s):
Contact e-mail: cuthbert@mit.edu
Research Area(s):
Combine HASS classwork with computer science by building the next generation of tools to enable research in history, anthropology, literature, global studies, CMS, and music and theater research in the new SHASS Digital Humanities lab. Work on textual and spatial analysis, and user interface especially encouraged.
"The Comédie-Française: Databases and Data Visualizations for an Eighteenth-Century French Theater"
Faculty Advisor: Jeff Ravel
Mentor(s):
Contact e-mail: ravel@mit.edu
Research Area(s):
The Comédie-Française Registers Project (cfregisters.org) consists of a database and set of visualizations that makes available detailed data on over 34,000 nightly theater performances at the preeminent Parisian theater during the French Enlightenment. In the first phase of this project we captured data based on nightly box office sales. We are now entering a new phase where we will capture data related to the troupe's daily expenditures and nightly casting choices for each play it performed. Tasks will include creation of an online data entry form linked to high resolution images of eighteenth-century documents, database additions and modifications, and the design of new visualizations that will allow users to explore the new data by itself, or jointly with the box office data already in the database. Working knowledge of French desirable, but not essential. The Comédie-Française: Databases and Data Visualizations for an Eighteenth-Century French Theater
"Teacher Practice Spaces for Equity Teaching Practices"
Faculty Advisor: Justin Reich
Mentor(s):
Contact e-mail: jreich@mit.edu
Research Area(s):
Great teachers know that learning requires practice, but ironically, teachers have very limited opportunities to develop their craft through practice and rehearsal. In the Teaching Systems Lab (tsl.mit.edu), we build online and blended practice spaces, inspired by games and simulations, where teachers can rehearse for and reflect on important decisions in teaching. We have a particular focus on working with computer science teachers to develop equity teaching practices- principles and strategies that help ensure that CS classes benefit all students, not just those historically represented in CS majors and careers.

In our work, we define equity teaching practices, co-design prototype practice spaces with teachers and teacher-educators, test our designs in lab-based playtests, and then field-test our refined products in the world through partnerships with school districts, colleges of education, and national CS professional development providers. Super UROPS may have the opportunity to refine and existing project, or start up a new prototype.

(Illustration from Swipe Right for CS, an app built by Junior Gabby Ballard and played by hundreds of CS teachers across the country through code.org-see tsl.mit.edu/gabby)
Teacher Practice Spaces for Equity Teaching Practices
"Scholarly Communication and Discovery"
Faculty Advisor: Heather Paxson
Mentor(s): Rodrigo Ochigame
Contact e-mail: paxson@mit.edu
Research Area(s):
Two possible development and design projects connected to Open Access publishing and online communication.

(1) Commons in a Box is a is a free software project that enables academic organizations to maintain a Commons site, where members discuss issues, collaborate, and share work. Using/retooling this platform, we can develop a Commons for humanistic social science (anthropology, history, science & technology studies). The project offers the opportunity to learn about scholarly publication information infrastructure (archiving, cataloguing, data metrics and management) and the chance to refine existing publication management systems towards envisioning new ones.

(2) Design novel algorithms and interfaces for the discovery of scholarly literature (search, recommendation, exploration) guided by principles of pluralism and critique, rather than metrics of "relevance" and "impact.” The aim is to help users become aware of and navigate a wider, more critical range of perspectives. Drawing on Open Access publications and metadata, we will modify and combine existing algorithms from graph theory to find patterns in the network of citations useful in identifying clusters of literatures. Such patterns, which do not indicate relations among literatures, must be complemented by curatorial methods. The project will develop extensions and plug-ins for free and open-source software used by libraries.
"Custom blocks as domain-specific languages for science modeling in StarLogo Nova"
Faculty Advisor: Eric Klopfer
Mentor(s): Daniel Wendel
Contact e-mail: djwendel@mit.edu
Research Area(s):
StarLogo Nova (www.slnova.org) is a blocks-based programming environment and simulation engine that allows teachers and students with no prior programming experience to use and create agent-based models of systems and phenomena that they are studying in science and math classes. Here is an example model of fish populations evolving in a pond, from one of our curriculum projects: Link

This project centers around the "custom blocks" feature in StarLogo Nova, in which users can create their own language primitives by parameterizing groups of other primitives, in a process that results in what are essentially libraries of functions, but feel like C++ macros. We want to study the use of this feature from a usability standpoint, as well as the pedagogical implications of these new primitives as they form what are essentially new domain-specific languages (DSLs). Do these DSLs aid students with no programming experience in understanding models built in StarLogo Nova? Or does the lack of consistency only add to potential confusion? Are there ways to support beneficial uses of these features and manage complexity well enough to mitigate potential negative effects?

The StarLogo Nova project also has many other sub-projects that could be turned into appropriate SuperUROP research projects in areas including computer graphics, programming language design, compilers, distributed systems, infrastructure and network architecture, security, and performance analysis and optimization. If you are interested in any of these other areas, email Daniel Wendel (djwendel@mit.edu) to arrange a meeting to discuss how your interests might intersect with other potential projects.
"Learning supports for first-time programmers in StarLogo Nova"
Faculty Advisor: Eric Klopfer
Mentor(s): Daniel Wendel
Contact e-mail: djwendel@mit.edu
Research Area(s):
StarLogo Nova (www.slnova.org) is a blocks-based programming environment and simulation engine that allows teachers and students with no prior programming experience to use and create agent-based models of systems and phenomena that they are studying in science and math classes. Here is an example model of fish populations evolving in a pond, from one of our curriculum projects: Link

This project aims to increase the learnability of StarLogo Nova through a design-based approach, combining heuristic analysis with user observation and direct feedback. Potential areas of development include integrated documentation, tutorial overlays, incremental unlocks, or even more sophisticated systems that involve community-supplied support. However, this is not only a design exercise; over 100,000 users are eagerly awaiting these additional supports, so ease of implementation and maintainability are just as important as hypothetical usability.

The StarLogo Nova project also has many other sub-projects that could be turned into appropriate SuperUROP research projects in areas including computer graphics, programming language design, compilers, distributed systems, infrastructure and network architecture, security, and performance analysis and optimization. If you are interested in any of these other areas, email Daniel Wendel (djwendel@mit.edu) to arrange a meeting to discuss how your interests might intersect with other potential projects.
"Automating Narrative Variation"
Faculty Advisor: Nick Montfort
Mentor(s):
Contact e-mail: nickm@nickm.com
Research Area(s):
The project focuses on how different arrangements of abstract narrative elements (events and existents in the story world) can be automatically realized in language (initially English). A modified and more easily accessible form of the Curveship system is used for this purpose. Curveship 0.5 has been released as a Python 2 framework for parser-based interactive fiction with narrative variation. In this form it has been used in creative and research work, leading to several publications. The new version of Curveship in ES6 (planned as 1.0) now exists with only core functionality; it runs in the browser and when complete will allow students, authors, players, and researchers to see different configurations of narrative at the story and discourse level. It will include all the narrative variation functionality of the original, with more to be added, but with a streamlined interface that does not include an IF parser, which was never used in research or teaching. A SuperUROP would work collaboratively with Prof. Montfort, creator of Curveship and programmer of both versions, and perhaps with others, to complete Curveship 1.0 for a free software release and to use it in research on narrative.

Total: 9

Technical questions related to this site? Please contact Search-Master@superurop-apply.mit.edu