Tackling Underrepresentation with Data Science

On February 24 at 12pm (EST), please join us for “Tackling Underrepresentation with Data Science.” The event will feature three lightning talks by professors and researchers in UNC-Chapel Hill’s academic community, centered around how data science is used across different disciplines to address underrepresentation. Speakers will include Deen Freelon, Andrés Hincapié, and Matt Jansen.

These talks will be followed by a guided panel, an opportunity for questions and answers with the speakers, and a discussion with the data science community at UNC-Chapel Hill, where we'll collaboratively examine the resources that enable researchers from a variety of disciplines to tackle underrepresentation through data science. Please register at the link below:

Speakers:

Our speakers for this month are as follows:

• Deen Freelon

  Associate Professor, Hussman School of Journalism and Media

  Quantitative social media research has traditionally been conducted from what might be called a platform-centric view, wherein researchers sample, collect, and analyzed data based on one or more topic- or user-specific keywords. Such studies have yielded many valuable insights, but they convey little about individual users’ tailored social media environments—what Freelon calls the user-eye view. Studies that investigate social media from a user-eye view tend to be rare because of the expense involved and a limited number of suitable tools. “Analyzing social media from a user-eye view with PIEGraph” introduces PIEGraph, a novel system for user-eye view research that offers key advantages over existing systems. PIEGraph is lightweight, scalable, open-source, OS-independent, and collects data viewable from mobile and desktop interfaces directly from APIs, incorporating an extensible tagging taxonomy that allows for straightforward classification of a wide range of political, social, and cultural phenomena. The presentation will focus on how our research team is using PIEGraph to examine users’ potential levels of exposure to high- and low-quality information sources across the ideological spectrum.

• Andrés Hincapié

  Assistant Professor, Economics

  Entrepreneurship in Black and White: Life Cycle Differences in Entrepreneurship between Black and White Males” uses a life cycle model of occupational choice to study various mechanisms explaining the racial gap in entrepreneurship. In the model, which is organized separately by race using data from the Panel Study of Income Dynamics, individuals can choose whether to work and whether to be paid- or self-employed — choices that are affected by their endowment of wealth and human capital, as well as by the process through which they transform these resources into earnings, a process which includes borrowing constraints. Additionally, their decisions are also influenced by their risk preferences and their non-pecuniary preferences for being self-employed. The estimates suggest that the main sources explaining the gap in entrepreneurship between black and white males are the returns to capital and the distribution of idea profitability.

• Matt Jansen

  Data Analysis Librarian, University Libraries (Digital Research Services)

  On the Books: Jim Crow and Algorithms of Resistance” is a collections as data and machine learning project of the University of North Carolina at Chapel Hill Libraries with the goal of discovering Jim Crow and racially-based legislation signed into law in North Carolina between Reconstruction and the Civil Rights Movement (1866/67-1967). This website lists and contextualizes North Carolina segregation laws for educators and researchers interested in Southern and African American History during the Jim Crow era. Jansen will be joined by project team members Amanda Henley, Lorin Bruckner, and Brianna Nuñez during the Q&A portion of the event to provide more context on the project overall.

Click below to register for the session now.

You can also sign up for our newsletter for more updates about Carolina Data Science Now and other data science opportunities.