Machine learning algorithms are as biased as the humans from which they emerged.
Many people view the computational machinery that collects research data as objective, however, the bias of the humans who invented and implemented the research lives on through the seemingly impartial machinery. This reality of imperfection in data collection perpetuates stereotypes and erasure across countless areas of research, including public health, gender equity and more.
Code^Shift is taking this data problem on.
Launched via Zoom on May 10, Code^Shift is a new multidisciplinary research lab for data for social justice, based at Texas A&M University. Code^Shift aims to develop knowledge, resources and community among data scientists and social justice-oriented scholars. This newly built community will, in turn, be useful for furthering social justice, collective healing, inclusion and transformation. Additionally, Code^Shift is one of four approved Thematic Data Science Labs proposals for the Texas A&M Institute of Data Science.
85 people - mostly graduate students and faculty - from six colleges at Texas A&M have already joined the List-Serv to stay connected. The goals of Code^Shift are community building, course training and social transformation through collective impact projects that lead to policy change.
Code^Shift’s codirectors are communication professor Dr. Srivi Ramasubramanian, computer science and engineering assistant professor Dr. Theodora Chaspari, computer science and engineering assistant professor Dr. Ruihong Huang and communication professor Dr. Lu Tang. This coalition of directors from various fields and perspectives allows for new solutions of numbers for social justice and inclusion.
Key in the methodology of the event was collaboration and connection, so breaking up the large crowd into smaller, more intimate groups was on the agenda. Six breakout rooms covered the following topics: LGTBQ+ Rights, Gender Equality, Health Disparities, Education & Literacy, Implicit Bias & Stereotypes and Racial Justice. Each room gave every person a chance to introduce their research, present their ideas and discuss their reflections on CODE^SHIFT.
Code^Shift is just beginning. Upcoming events include the Discover-Text Workshop on May 24, the Health Equity Panel on June 7 and the “Watch & Talk: Coded Bias” event in which participants will watch a documentary and discuss the film afterward. Before all of the Fall 2021 events ensue, Code^Shift’s full proposal is due on July 2.