PublicationCancer registry coding via hybrid neural symbolic systems in the cross-hospital setting(在新分頁開啟)
2025MedInfoSpeech De-identification of Chinese, English and Minnan: Effectiveness of Chinese-based LLM Model and ASR(在新分頁開啟)
2025FSRTaipei, TaiwanImproving Low-Resource Speech Recognition with Whisper-MoE and Synthetic Data Augmentation: A Case Study on Hakka(在新分頁開啟)
2025AACRTainan, TaiwanRepresent Cancer Data as Knowledge Graph along with Large Language Models to Identify Recurrence Patterns