Smishing Detection
- Built a smishing detection model using DistilBERT, TF-IDF, and Named Entity Recognition (NER) to identify malicious messages.
- Tackled real-world challenges like imbalanced datasets, improving model accuracy and reliability through strategic data handling.
- Independently managed key tasks while maintaining clear communication and collaboration within a team environment.
- Applied ethical and privacy-conscious practices throughout the project, with a strong focus on secure and responsible AI use.
Environmental Resource Consumption and Service Demand in Melbourne
- Led a data-driven project focused on analyzing the relationship between customer service request trends, resource consumption, and planned capital works in the City of Melbourne.
- Cleaned, integrated, and preprocessed multiple datasets including customer service requests, environmental resource usage, and capital works data.
- Conducted exploratory data analysis (EDA) to identify patterns, seasonal trends, and operational inefficiencies.
- Performed temporal and correlation analyses to uncover relationships between service demand, resource consumption, and greenhouse gas emissions.