I recently defended my Ph.D. in Computer Science at the University of Southern California (USC), where I was a member of the Morality and Language Laboratory (MOLA) advised by Prof. Morteza Dehghani.

I strive to understand machine learning pipelines in the context of the larger socio-technical system they’re embedded into. My research has predominantly focused on mitigating problematic content in the digital realm. I’ve addressed diverse technical challenges to reduce online problematic content, whether it stems from human users or machine-generated models. A core component of my work is acknowledging the societal roots of this issue by advocating for and pursuing approaches informed by social science theories.

Overview of My Research on Mitigating Problematic Content in Digital Sphere Overview of My Research on Mitigating Problematic Content in Digital Sphere

During my Ph.D. at the USC, I interned twice for Snap Inc. Over the two summers at Snap Inc., I made key contributions to enhancing their machine learning advertising technology. In 2023, I designed, developed, and deployed a language model integrated into Snap’s identity framework to learn user patterns on advertiser websites, achieving a 3% boost in precision that was adopted into production. The prior year in 2022, I constructed a machine learning pipeline to accurately categorize advertisement website content based on a taxonomy, attaining an F1 score over 0.8. This system was deployed for tag suggestions, streamlining the annotation process. These internships provided hands-on experience applying machine learning to improve Snap’s user tracking and content analysis capabilities for their advertising platforms.

Prior to MOLA, I finished my B.Sc. in computer engineering (majoring in software) at the Department of Electrical and Computer Engineering, University of Tehran in August 2019. I tried to use my bachelor’s to build a solid bedrock for my future research. So in addition to taking many optional graduate-level courses on math and computer science, I started by investigating the Call Data Records of Iran’s top mobile operators at the Data Science Laboratory at the University of Tehran as a keen freshman. Then, trying to stay at the edge of research, I spent the 2017 summer at Fraunhofer IDMT in Germany researching on the applications of Deep Learning to Music Information Retrieval. After Fraunhofer, following my passion to better understand human behavior, I spent a year as a Research Assistant at the Cognitive Systems Laboratory at the University of Tehran where I compeleted my thesis on utilizing attention mechansim in neural networks and won the Best Thesis Award for it.

During my bachelors, I was also affiliated with RoboCup Asia-Pacific (RCAP) 2018, a super-regional headquarter representing the RoboCup Federation, coordinating RoboCup activities in the Asia Pacific region. Before joining University of Tehran, I was a high school student at the Allameh Helli Highschool which is affiliated with NODET.