On June 15th, the University of Washington’s Data Science for Social Good (DSSG) summer program began with eight fellows working on two project teams in collaboration with data scientists from the eScience Institute and project leads from academic institutions and nonprofit organizations. Student fellows joined the program from colleges and universities around the country. Their diverse areas of study include media and information, public policy, biomedical informatics, energy and resources, public health, quantitative methods in the social sciences, and physics.
Project leads from academia, public agencies, nonprofits and industry guide teams through the ten-week program, run by the eScience Institute. Fellows participate in tutorials and workshops covering quantitative and qualitative data science tools, methods and approaches, such as Git and GitHub, coding standards and documentation, machine learning, ethics, human-centered design, and reproducible science. Each team also participates in systematic and sustained engagement with an array of related stakeholders. This year, the program will take place remotely for the first time and will be partially supported by an Advancing Curiosity Grant from the Micron Foundation.
This year’s program supports two extremely timely and relevant projects.
“Identifying Coronavirus Disinformation Risk on News Websites” is led by Maggie Engler, Lead Data Scientist, and Lucas Wright, Senior Researcher, at the nonprofit organization Global Disinformation Index (GDI). GDI combats disinformation by helping online advertisers to reduce the appearance of their ads on websites that spread disinformation, with a goal of shrinking funding sources to organizations that misinform. This project will create an automated system to classify news articles according to their risk of containing disinformation about the coronavirus, thereby enabling advertisers to prevent their ads from appearing on websites that contain erroneous news stories.
The team also includes fellows George Hope Chidziwisano, Richa Gupta, Kseniya Husak, Maya Luetke and data scientists Vaughn Iverson and Noah Benson. Fellows will collaborate with GDI researchers to combine data from diverse sources, perform natural language processing on tens of thousands of news articles, highlight misinformation about the coronavirus, and document their code and test results.
“Detection of Vote Dilution: New tools and methods for protecting voting rights” updates a software package called ‘eiCompare R’ with an improved methodology for estimating the race and ethnicity of voters, in order to more accurately identify communities with “racially polarized” voting patterns. Demonstrating such patterns of disparate voting among whites and minorities allows voters to challenge district boundaries in court under Section 2 of the Voting Rights Act. The project builds on existing software to modernize ecological inference (EI) analysis, a statistical method used by social scientists that relies on imprecise census data to estimate voter race and ethnicity.
The project is led by Matt A. Barreto, Professor of Political Science and Chicana/o Studies, and Faculty Director of the Voting Rights Project at University of California, Los Angeles; and Loren Collingwood, Associate Professor in the Department of Political Science at University of California, Riverside. The team includes fellows Juandalyn Burke, Ari Decter-Frain, Hikari Murayama and Pratik Sachdeva, and data scientists Scott Henderson and Spencer Wood. The fellows will contribute to a range of methodological, programming, and statistical advancements to the EI models in eiCompare to improve accuracy in capturing racial voting patterns, in support of further legal efforts.
Visit the eScience website to learn more about this year’s DSSG projects and participants. The program will conclude with project presentations to a public audience on Wednesday, Aug. 19 from 1 to 3 p.m. Additional information will be available later this summer.