SuperUROP Projects

"A global, machine learning-based analysis of the relationship between policy commitments and greenhouse gas emissions" Faculty Advisor: Evan Lieberman Mentor(s): Contact e-mail: evanlieb@mit.edu Research Area(s):
Solving the climate crisis will require that the commitments of leading political and economic actors around the world actually translate into substantial reductions in greenhouse gas emissions. Which forward-looking promises actually presage the actions required to attain such results? The new, freely available Climate Trace database provides a remarkable tool for measuring emissions trends at a very high-resolution around the world. This super-UROP research project will seek to amass a large global corpus of policy documents produced by firms and governments, and will attempt to identify the language patterns in those documents that predict GHG emissions trends. It will also seek to detect language patterns that promise but do not deliver on GHG reductions. The first step of the project will involve the recovery of a large corpus of documents geo-located as associated with a sample of baseline emitters in a range of industries; Second, we will train and implement natural language processing models for classifying different types of policy intentions and commitments within those documents; and finally, we will use various data analytic tools to identify the most promising language-based policies for emissions reductions, helping to answer the key research questions that motivate the project.
"Tracing trade policies of the US over the last 250 years using machine learning" Faculty Advisor: In Song Kim Mentor(s): Contact e-mail: insong@mit.edu Research Area(s):
Our SuperUROP project is diving into an exciting challenge: mapping out the U.S. trade policies from way back in 1789 all the way to today. We're using a mix of advanced machine learning methods to sift through hundreds of years of legislative bills, aiming to uncover how trade policies, say for something as everyday as sugar, have shifted over time. This deep dive isn't just about tracking changes; it's about understanding the intricate political relations behind these policies and how they've evolved to address increasingly complex interests. This is a fantastic chance for students interested in applying their machine learning know-how from computer science to dig into the rich field of international trade history. You'll get to explore how detailed trade policies are crafted and how they mirror the political and economic issues of their times. We're looking for students who are excited about blending tech skills with historical insights to join us. You'll sharpen your data science techniques while gaining a unique perspective on the global trade dynamics that shape our world. This project is more than an academic exercise; it's an opportunity to explore the interconnections of politics, economics, and history through the powerful lens of machine learning. Come join us and help uncover the stories behind centuries of trade policies.
"Investigating political activities of US companies using machine learning methods" Faculty Advisor: In Song Kim Mentor(s): Contact e-mail: insong@mit.edu Research Area(s):
Our team has developed an innovative database that maps out the lobbying and campaign donation activities of US companies over the past quarter-century. This unique, high-dimensional dataset opens up a window into the complex interplay between individuals, politicians, companies, and entire industries. Through our SuperUROP project, students are invited to embark on a deep dive into these connections, employing advanced machine learning techniques to unearth insights that could significantly impact our understanding of major societal challenges. From environmental sustainability and the ethical dimensions of artificial intelligence to the intricacies of global investments and international relations, this project offers students a rich platform for exploration and discovery. By participating, students will not only refine their technical and analytical skills but also contribute to meaningful change by shedding light on how corporate activities shape public policy and societal outcomes. We are looking for students who are eager to leverage machine learning to make sense of complex data and are passionate about applying their findings to address some of the most pressing issues of our time. Join us in this endeavor, through the CS+HASS Advanced Undergraduate Research Program, to not just learn, but to illuminate the pathways through which business and politics intersect and influence our world.

Total: 3