NLP & Machine Learning Research, Ashesi University
Performed NLP research involving model evaluation on real-world noisy text, reaching high F-1 performance. Built and standardized evaluation workflows with automated validation and error logging to improve consistency and reduce manual effort. Cleaned preprocessing pipelines to enhance model accuracy and reliability for downstream analysis. • Built sentiment analysis models (CNN and RNN) for text classification. • Standardized NLP evaluation workflows with automated pipeline checks. • Improved output consistency through error logging during validation. • Achieved 90% F-1 by cleaning preprocessing pipelines.