PATRICK VAN KESSEL
Data Science Associate, Pew Research Center
MONDAY, MARCH 20th, 2016
12:00 – 1:30pm | MIT E53-482 | 70 Memorial Dr., Cambridge, MA

RSVP here by Friday, March 17th.

Description: Polarization in Congress has been increasing since the 1970s, and more recent decades have seen the growth of similar divisions among the American public. While political scientists have long thought that politicians’ communications about their positions and priorities may mediate these two trends, few have systematically analyzed this rhetoric on a large scale. To shed light on the communication thought to link public and elite polarization, Pew Research Center’s Data Labs collected, cleaned, and analyzed more than 200,000 press releases and Facebook posts issued by members of the 114th Congress.

Using crowd-sourcing and supervised machine learning models, Data Labs quantified how often legislators use targeted political criticisms to “go negative” in their outreach to the public, and found that their followers reward them with increased social media attention for doing so. Patrick van Kessel, one of the project’s lead researchers, will discuss the implications of these findings, explore the technical and methodological challenges presented by this project, and describe the data science tools and approaches that were used to overcome them.

Patrick van Kessel is a Data Science Associate at Pew Research Center, where he conducts computational social science research that involves machine learning, natural language processing, data systems engineering, and large-scale web-based data collection. As a professional data scientist and research methodologist, he has applied these tools and methods across a variety of research areas, including political communication, criminal justice, public health, international security, early childhood education, and positive psychology.

Pew Research Center’s Data Labs uses computational methods to complement and expand on the Center’s existing research agenda. The team collects text, network and behavioral datasets; uses innovative computational techniques and empirical strategies for analysis; and generates original research. Data Labs also explores the limitations of these data and methods and works toward establishing standards for use and analysis.

__

The Data Science to Solve Social Problems series features practitioners who are applying data science techniques to real world social problems. This series aims to promote dialogue and collaboration between social scientists and data analysts / engineers working on innovative projects. For more information on speakers and to get involved, contact Soubhik Barari at sbarari@mit.edu.