🤖 AI Summary
Historical documents are often rendered partially illegible due to physical degradation, posing significant challenges—particularly in recovering proper nouns that rely heavily on external contextual knowledge. This work proposes a novel framework that integrates implicit knowledge from large language models with explicit historical context retrieved from external knowledge bases, leveraging retrieval-augmented generation (RAG) for the joint restoration of both general characters and named entities. By incorporating context-aware reasoning to effectively fuse domain-specific knowledge, the method substantially outperforms existing baselines on Korean historical documents, achieving marked improvements in both character-level and named entity recovery accuracy. The approach has also been endorsed by domain experts as a practical tool for historical text analysis.
📝 Abstract
Historical documents act as invaluable knowledge archives but often suffer from illegibility due to physical deterioration and damage. While existing restoration methods based on masked language modeling effectively utilize local context, they struggle to restore named entities that require external historical knowledge. To address this limitation, we introduce a novel framework for historical document restoration that leverages large language models with retrieval-augmented generation (RAG). By combining the implicit knowledge of pre-trained LLMs with explicitly retrieved external context, our model ARI effectively mitigates the challenge of inferring context-dependent proper nouns. Extensive experiments on Korean historical documents demonstrate that our approach significantly outperforms baselines, achieving substantial gains in restoring both general characters and named entities. Furthermore, comprehensive evaluations including expert assessments confirm that ARI serves as a practical tool for domain experts, promising to accelerate the analysis of historical records.