透過溝通學習與知識結晶化提升低資源語言能力：以臺灣臺語為例;Enhancing Low-Resource Language Capabilities via Communicative Learning and Knowledge Crystallization A Case Study on Taiwanese Hokkien

NCUIR > College of Electrical Engineering & Computer Science > International Graduate Program in Artificial Intelligence > Electronic Thesis & Dissertation > Item 987654321/99167

Please use this identifier to cite or link to this item: https://ir.lib.ncu.edu.tw/handle/987654321/99167

Title:	透過溝通學習與知識結晶化提升低資源語言能力：以臺灣臺語為例;Enhancing Low-Resource Language Capabilities via Communicative Learning and Knowledge Crystallization A Case Study on Taiwanese Hokkien
Authors:	朱芸歆;Chu, Yun Hsin
Contributors:	人工智慧國際碩士學位學程
Keywords:	低資源語言;溝通式學習;知識結晶化;互動式學習;大型語言模型;Low-Resource Language;Communicative Learning;Knowledge Crystallization;Interactive Learning;Large Language Model
Date:	2026-01-16
Issue Date:	2026-03-06 18:15:15 (UTC+8)
Publisher:	國立中央大學
Abstract:	為了確保低資源語言適應過程中的知識可攜性與運作成本效益，本研究旨在解決大型語言模型在長程互動中面臨的指令碎片化、高昂推論成本以及規則遵循度下降等挑戰。本研究提出了一種「知識結晶化」機制以實現知識的可攜性，透過將短暫的對話互動提煉為高純度的語言規則，成功將閉源大型模型的語言專業知識遷移至更具成本效益的模型中。針對平行語料稀缺的「台灣台語」，本研究提出了一套結合溝通式學習與知識結晶化機制的創新「跨模型師生架構」。我們採用 Gemini 2.5 Pro 擔任「教師模型」，負責生成合成資料，並透過互動式對話指導 GPT-5.1 與 DeepSeek-v3 等「學生模型」。為解決直接堆疊原始對話所導致的指令碎片化與高推論成本，我們引入了「知識結晶化」機制。此機制的設計初衷並非僅為緩解記憶體限制，而是旨在透過「後設認知反思」，將稍縱即逝的互動過程提煉為高純度的語言規則。實驗結果顯示，在錯誤修正方面，互動式策略的表現顯著優於被動示範。值得注意的是，DeepSeek-v3 展現了卓越的適應性，能有效激活其潛在的多語言能力，在無需更新參數的情況下即逼近當前最先進LLM的水準。效率分析指出互動成效在第5輪達到峰值，實證了結晶化機制對於維持高資訊密度及優化 Token 消耗的必要性。本研究建立了一套具成本效益且免微調的範式，成功將通用模型與低資源語言對齊，為台灣台語的數位保存做出了貢獻。;To ensure instructional stability and economic efficiency in Low-Resource Language (LRL) adaptation, this study addresses the challenges of instructional fragmentation, high inference costs, and reduced rule-adherence during long-form LLM interactions. This study proposes a Knowledge Crystallization mechanism to achieve knowledge portability, distilling ephemeral interactions into high-purity rules to transfer linguistic expertise from closed-source giants to more cost-effective models. Focusing on Taiwanese Hokkien, a language characterized by limited parallel corpora, this study proposes a novel Cross-Model Teacher-Student framework that integrates communicative learning with a knowledge crystallization mechanism. We employ Gemini 2.5 Pro as the Teacher model to generate synthetic data and guide Student models (GPT-5.1 and DeepSeek-v3) through interactive dialogue. To address the instructional fragmentation and high inference costs associated with raw dialogue stacking, we introduce a "Knowledge Crystallization" mechanism. This process is designed not merely to accommodate memory constraints, but to distill ephemeral interactions into high-purity linguistic rules via metacognitive reflection. Experimental results demonstrate that interactive strategies significantly outperform passive demonstration in error correction. Notably, the DeepSeek-v3 model exhibits exceptional adaptability, effectively activating latent multilingual capabilities to approach SOTA levels without parameter updates. An efficiency analysis identifies a peak at 5 interaction turns, empirically validating that crystallization is essential for maintaining high information density and optimizing token consumption. This work establishes a cost-effective, fine-tuning-free paradigm for aligning general-purpose models with LRLs, contributing to the digital preservation of Taiwanese Hokkien.
Appears in Collections:	[ International Graduate Program in Artificial Intelligence ] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	85	View/Open

社群 sharing

Loading...