Project-09-Pioneering Turkish Emotion Analysis with LLMs
Executed one of the first studies to fine-tune pre-trained Large Language Models specifically for Turkish emotion analysis using curated, domain-specific datasets. By leveraging novel transfer learning techniques, the research achieved state-of-the-art accuracy scores, surpassing traditional baselines. This work resulted in 3 published papers and established a new benchmark methodology for Turkish computational linguistics. (2017-2020)