문맥 인식 기반 PDF 키워드 마스킹을 위한 학습 시스템 설계 및 구현

초록

This paper proposes PDFMask, a context-aware keyword masking learning system designed to improve learning efficiency in digital PDF environments. The proposed system utilizes textual and positional information embedded in PDF documents to reconstruct the logical reading order and automatically identify key terms using morphological and syntactic analysis. A masking PDF suitable for learning is then generated by applying ratio-based sampling and visual annotation while preserving the original document layout. In addition, an asynchronous message queue-based architecture that separates the web server from the processing engine is adopted to enable efficient handling of large-scale PDF processing requests. The proposed system alleviates the loss of contextual information observed in flashcard-based learning and reduces the manual effort required in PDF annotation-based approaches.

키워드