Data Scientist — Protect and Care
- Developed algorithms and dashboards to support moderation and online safety.
- Identified levers for prioritizing content moderation, minimizing violating view by 85%.
- Prototyped pipelines and RShiny dashboards to monitor and detect high risk content on the platform.
Data Scientist — Data Labs
- Implemented deep learning models for unsupervised learning, computer vision, and natural language processing tasks.
- Engineered interpretable features for content based recommendation using multi-modal item embeddings (text, image, context).
- Designed and analyzed human computation experiments to study ranking algorithms effects on human performance using mixed effects models.
Data Scientist — Forward Deployed
- Consulted clients to better understand their modeling requirements.
- Developed and tested the first lookalikes algorithm for audience expansion tools,
NYU Global Institute of Public Health
Research Assistant — ChunaraLab
Supervisor: Dr.Rumi Chunara
- Analyzed over 50 million tweets to study the geographic and temporal characteristics of drug usage among twitter users in north america.
- Leveraged Amazon Mechanical Turk to build an active learning pipeline for heirarchical classification using SVMs to minimize costs for dataset creation.
Data Scientist — Sysomos Labs
- Prototyped advertisement optimization platform for offline targeted ads using multiarmed bandit algorithms.
- Improved clustering performance for twitter community detection on Sysomos customer products.
Liu J., Weitzman E., Chunara R.
Assessing Behavior Stage Progression from Social Media Data
Computer-Supported Cooperative Work and Social Computing 2017, Portland, OR, USA
Rehman NA, Liu J., Chunara R.
Using Propensity Score Matching to Understand the Relationship Between Online Health Information Sources and Vaccination Sentiment.
AAAI Spring Symposium. 2016, Stanford University Palo Alto, CA, USA