top of page
Create Your First Project
Start adding your projects to your portfolio. Click on "Manage Projects" to get started
StackOverflow Data Pipeline
Project type
Machine Learning and Natural Language Processing for Hate Speech Detection
Date
May 2025
Location
Deakin University, Victoria, Australia
Link
Built an end-to-end pipeline to process StackOverflow XML data, converting it into structured datasets, performing feature engineering, and training multiple ML models (Logistic Regression, Random Forest, XGBoost, LightGBM) to predict accepted answers and post views. Applied SMOTE, ADASYN, PCA, and SHAP for imbalance handling, dimensionality reduction, and explainability.
bottom of page