Sentiment Annotation for Emotion Classification (Academic Project)
Built and applied a sentiment analysis model to classify emotions in over 12,000 customer reviews from Tiki Books. Used PhoBERT for emotion classification and data reliability evaluation as part of a scientific research project. Focused on systematically annotating and categorizing textual data for model training. • Conducted automated and manual emotion labeling of customer feedback. • Evaluated and verified data label quality for sentiment analysis. • Utilized Python, Pandas, and NLP libraries for annotation workflow. • Ensured consistent labeling criteria and reliability across dataset.