Who I am
PhD student in Informatics
At Pennsylvania State University, I work on privacy, web-scale text analysis, and responsible AI questions.
PhD Student in Informatics - Pennsylvania State University
Studying how privacy norms evolve from the conventional web to agentic AI
PhD student in Informatics at Penn State studying privacy policy analysis, sectoral web classification, and disclosure behavior in AI agents.
Overview
Overview
I study privacy disclosures in two settings: the conventional web and emerging agentic systems. The site centers on the datasets, analyses, and evaluation work that connect those settings.
What I study
Privacy norms across two environments
My work tracks privacy disclosures in websites and asks how those norms shift when AI agents act on a user’s behalf.
Where to start
SoACer, PrivaSeer, and agentic disclosure
These projects show the main arc of the site: sector discovery, privacy analysis, and agentic AI evaluation.
- 3M+
- privacy policies in PrivaSeer
- 195,495
- websites in the SoAC corpus
- 59,590
- sector-tagged privacy policies analyzed
Research Map
Research Map
Three themes organize the current work.
Sectoral web classification
I build datasets and models that recover service sector from website text, making sector context usable for downstream analysis.
Privacy policy analysis at scale
I study how disclosures vary across industries and over time, with attention to transparency, vagueness, and sector-specific norms.
Agentic AI disclosure
I examine what AI agents reveal during tool use and how model behavior and schema design shape oversharing risk.
Selected Work
Selected Work
A few projects that show the main arc of the research.
SoACer and the SoAC Corpus
Problem
Sector context is hard to recover from noisy and heterogeneous website content.
Contribution
SoACer and the SoAC Corpus provide a web-scale classification pipeline and dataset for sector-aware analysis.
Why it matters
They make privacy and governance studies more useful by preserving service context instead of flattening the web.
PrivaSeer and sectoral privacy analysis
Problem
Privacy disclosures are difficult to compare across industries and time at scale.
Contribution
My work uses PrivaSeer and sector-aware analyses to study convergence, divergence, and opacity in privacy policies.
Why it matters
This creates empirical grounding for transparency, governance, and data-handling research.
Agentic privacy disclosure
Problem
AI agents may disclose user information too similarly across service contexts when they use tools.
Contribution
My dissertation work treats runtime tool calls as a key disclosure moment and studies both model behavior and schema design.
Why it matters
It extends privacy analysis from the conventional web to agentic systems.
Selected Publications
Selected Publications
Selected papers relevant to the homepage. Full publication details are on Google Scholar and the CV.
2025
SoACer: Sector-Based Corpus and LLM-Based Framework for Sectoral Website Classification
Builds a dataset and classification pipeline for sector-aware web analysis.
2025
The PrivaSeer Project: Large-Scale Resources for Analysis of Privacy Policy Text
Provides infrastructure for large-scale transparency and privacy-policy analysis.
2022
Generative Adversarial Learning with Negative Data Augmentation for Semi-Supervised Text Classification
Earlier work on robust NLP under low-label conditions.
Academic Snapshot
Academic Snapshot
Current work
Graduate Research Assistant
Work on PrivaSeer, sector-aware privacy analysis, and large-scale empirical studies of disclosure behavior.
PhD in Informatics
Dissertation on the evolution of sectoral privacy norms from the conventional web to the agentic web.
Previous research and service
Research Intern
Applied NLP methods to qualitative corpora and translated findings into policy-relevant insights for public stakeholders.
Graduate Research Assistant
Studied fairness and robustness in semi-supervised text classification and built the work that became my master's thesis.
Reviewer
Reviewed work on trustworthy NLP and responsible model behavior.
Graduate Student Panelist
Spoke with students about research careers and graduate study.
Collaboration and Contact
Collaboration and Contact
I welcome conversations about privacy, AI governance, responsible NLP, and agentic systems, especially around research collaboration, datasets and evaluations, contributor work, and research opportunities.
- Research collaboration
- Datasets, evaluations, and contributor work
- Academic and applied research opportunities