Aim High, Stay Private
Harvard University · Boston, MA
I presented utility trade off of differentially private synthetic data generators for the LEMURS dataset, a behavioural health study done at the UVM.
Read the LinkedIn post →Privacy · Machine Learning · Data Governance
Mohsen
Ghasemizade
PhD Candidate, Computer Science — University of Vermont
MassMutual Research Fellow · Computational Ethics Lab
Research
Differential Privacy & Synthetic Data
Privacy-Preserving ML & LLMs
Policy & Regulatory Compliance
Computational Social Science
Selected Work
Publications & projects
Taste for Privacy: How Context, Identity, and Lived-Experience Shape Information Sharing Preferences
A study of 2,912 survey responses from 782 college students mapping how comfort sharing personal data shifts across 17 institutional contexts, and how demographics and lived experience reshape institutional trust.
Read paper →Aim High, Stay Private: Differentially Private Synthetic Data Enables Public Release of Behavioral Health Information with High Utility
A differentially private pipeline that anonymizes weekly Oura-ring and survey data from 600+ college students, enabling public release of behavioral-health insights while protecting individual privacy.
Read paper →Developing a Hierarchical Model for Unraveling Conspiracy Theories
A BERT/RoBERTa text classifier (87% F1) for identifying sentiment in conspiracy-theory discourse, with data collected via large-scale web scraping.
Read paper →Making Local Government Contracts Legible: A Computational Pipeline for Classifying and Mapping Intergovernmental Service Agreements
An end-to-end pipeline — neural OCR, LLM-based classification, and network modeling — applied to 21,000+ Iowa 28E agreements (96% extraction accuracy, 82% weighted F1), revealing a stratified buyer–seller structure across government types.
Privacy Policy Compliance Scoring System
A scoring system using a fine-tuned LLMs model to evaluate how tech-company privacy policies have evolved against regulation from the 1980 OECD guidelines through GDPR and CCPA.
Talks & Presentations
Membership Inference Attack on Tabular Data
Northeastern University · Boston, MA
We presented our poster for the new membership inference attack on tabular data using gaussian distribution to fit the attack model.
Read the LinkedIn post →
Data Exploration with Python
EUniversity of Vermont · Burlington, VT
I presented Data Exploration with Python to talented high school students from all over the state of Vermont, who were residing on the UVM campus, as part of The Governor's Institutes of Vermont - Math group.
Read the LinkedIn post →About
I’m a computer science PhD student at the University of Vermont and a MassMutual Research Fellow in the Computational Ethics Lab. My work sits at the intersection of differential privacy, machine learning, and data governance — designing systems that let researchers learn from sensitive data without compromising the people it describes.
I earned my MS in computer science at UVM, with a thesis on a computational, genealogical approach to conspiracy theories, and a BS in electrical engineering from Shiraz University. Recent work spans differentially private synthetic health data, privacy-policy compliance analysis with LLMs, and large-scale text and network studies of public-sector data.
Outside research, I serve on the board of Local Motion, advancing safe streets and active transportation across Vermont — and I spend a lot of time on a bike.
Experience & Education
- 2026 —R&D Intern · MassMutual Differential privacy for secure sharing of longitudinal health data.
- 2026 —Board Member · Local Motion Safe-streets and active-transportation policy lead.
- 2022 —Graduate Research Assistant · Computational Ethics Lab, UVM
- ’22–’24Graduate Teaching Assistant · UVM Co-instructed undergraduate Python; mentored 150+ students.
- 2024 —PhD, Computer Science · University of Vermont MassMutual Research Fellow · Chernoff Fellowship.
- 2024MS, Computer Science · University of Vermont
- 2020BS, Electrical Engineering · Shiraz University