My research sits at the intersection of computational methods and social inquiry. I am interested in how machine learning, simulation, and data analysis can be applied not merely to process data efficiently, but to understand social phenomena — bias, conflict, trade dynamics, group behavior — in ways that are rigorous, interpretable, and meaningful. I approach research with attention to both technical validity and social relevance, and I am drawn to questions that require engaging seriously with both the computational and the social dimension of a problem.
Research Interests
1. Computational Social Science
Applying quantitative and computational methods to study patterns in human behavior, social structures, and collective dynamics. I am interested in how large-scale social phenomena can be analyzed empirically using machine learning and statistical modeling, and in the epistemological questions this raises: what can a computational model capture, and what does it necessarily leave out? My work explores a range of methods applied to social questions, aiming for approaches that are both technically sound and socially grounded.
2. Natural Language Processing & Text Analysis
Using NLP methods to extract social meaning from textual data. I am interested in bias detection, sentiment analysis, and the identification of cultural and structural patterns embedded in language. My work on gender bias in literary book reviews drew on TF-IDF weighting, sentiment scoring, and statistical hypothesis testing to examine how review language differs across author gender. I approach text analysis with attention to both its analytical reach and its interpretive limits.
3. Agent-Based Modeling & Social Simulation
Designing computational models in which individual agents follow local rules and interact to produce emergent collective behavior. I have used the Mesa library in Python to build a spatial conflict model examining how the density and distribution of shared Points of Interest shape inter-group dynamics. Agent-based simulation complements statistical inference — it is particularly valuable for studying non-linear, path-dependent social processes and for testing theoretical assumptions about collective behavior.
4. Data Visualization & Visual Analytics
Making complex, multidimensional socio-technical data legible through interactive and dynamic visualization. My work on global energy trade flows involved processing large structured datasets and constructing navigable visual tools for exploring cross-temporal trade relationships. Visualization is not merely a communication medium — it is a form of analysis. The choices made in representing data shape what patterns can be seen and what questions can be asked.
5. Responsible AI & Socio-Technical Systems
Examining the fairness, interpretability, and social implications of data-driven systems. I am interested in how AI and algorithmic systems encode social values and produce social effects — and how researchers can engage with this responsibility more deliberately. This concern runs across my work on bias detection, platform design, and simulation, and reflects my broader commitment to research that is aware of its social context.
Methods & Tools
Data Analysis & Engineering
Python · Pandas · NumPy · SQL · Exploratory Data Analysis · Data Cleaning & Preprocessing · Feature Engineering · Time Series Analysis · Spatial Interpolation (KNN, IDW, Fourier)
Machine Learning & Modeling
Scikit-learn · PyTorch · Linear Regression · Logistic Regression · Decision Trees · Random Forest · K-Means Clustering · PCA · Model Evaluation (RMSE, R²)
Natural Language Processing
TextBlob · TF-IDF · Sentiment Analysis · Statistical Hypothesis Testing (t-test, Mann-Whitney U)
Simulation
Agent-Based Modeling · Mesa (Python) · Spatial Simulation Design
Visualization
Matplotlib · Interactive Dashboards · Visual Analytics Design
Open Science
I am committed to transparent and reproducible research practices. Where project conditions allow, I make associated code and data available for inspection and reuse. Openness in methodology and data is a foundational commitment for credible computational social science, and I aim to build this practice consistently across my work.