The Old vs. New Debate in Named Entity Recognition: A Comprehensive Comparative Study of Encoder Models and Generative Large Language Models

Introduction / Main Floor: The Old vs. New Debate in Named Entity Recognition: A Comprehensive Comparative Study of Encoder Models and Generative Large Language Models

A bachelor's thesis-level systematic comparative study that deeply compares the differences in performance, efficiency, and robustness between the traditional encoder architecture (DeBERTa) and the generative large language model fine-tuned with LoRA (Qwen3.5) on named entity recognition tasks, providing empirical evidence for model selection in real-world application scenarios.