Sentiment Analysis of Imbalanced Dataset Through Data Augmentation and Generative Annotation Using DistilBERT and Low‐Rank Fine‐Tuning
Experimental results on the Twitter US Airline Sentiment dataset demonstrate that the proposed framework achieves strong classification performance while maintaining low training complexity, highlighting the effectiveness of combining Large Language Model (LLM) based data augmentation with parameter‐efficient transform...