Three graduate students from the University of Arkansas at Little Rock were selected for a prestigious workshop designed to prepare graduate students for careers in the computer science industry.

Adewale Obadimu, a computer science doctoral student, and graduate students Oluwaseun Johnson and Uche Umoga participated in the 2020 CRA URMD Grad Cohort Workshop March 5-7, 2020 in Austin, Texas.

The three-day event was hosted by the Computing Research Association (CRA), an organization dedicated to bringing industry, government and academia together to support research and advanced education in computing. The third annual CRA Grad Cohort for Underrepresented Minorities and Persons with Disabilities (URMD) Workshop attracted students from schools across the United States and Canada. CRA covered the travel expenses for all selected students that collectively represent a diverse set of computing-related research areas and institutions.

Participants attended a wide variety of professional development sessions on topics including academic workflow, career options, and mental wellbeing. The focus was to learn from one another leaving not only with more knowledge, but also a greater professional network.

“I enjoyed networking with graduate students from other institutions and professionals from IBM, Google, Microsoft, and Bloomberg,” said Johnson. “I had an opportunity to discuss my research with other participants and hear talks on networking, building your professional persona, summer internships, balancing graduate school and personal life, strategies for human to human interaction, and entrepreneurship.”

Obadimu presented his poster titled “Identifying Latent Toxic Features on YouTube Using Non-negative Matrix Factorization.” The contribution of this work is to help understand the relationship between the impact of negativity, or toxicity, of a video on the comments, to enable us to predict the likely toxicity of a commenter based on their past history, and to allow us to determine what kind of comments a video will generate based on prior toxicity matrix.

“During the poster session, I reviewed other research projects and gained some valuable insights into the do’s and don’ts of effectively convey poster information,” Umoga said.

Obadimu, Johnson, and Umoga are the lead developers of the Blogtrackers tool that helps COSMOS conduct discourse analysis and event analysis on blogs, including assessment of misinformation and fake news.