초록

용어보기논문에서 용어와 풀이말을 자동 추출한 결과로, 시범 서비스 중입니다. AI-Helper

Abstract ▼ AI-Helper

Recent rapid rise in plagiarism has become a serious social issue and has prompted increasing research on copy detection. Unfortunately, current plagiarism detection systems are concerned with detecting and collecting documents but failed to protect the original documents. Since authors usually do n...

Recent rapid rise in plagiarism has become a serious social issue and has prompted increasing research on copy detection. Unfortunately, current plagiarism detection systems are concerned with detecting and collecting documents but failed to protect the original documents. Since authors usually do not want their work to be collected for plagiarism inspection without their permission, plagiarism detection and source document protection are in conflict with each other. We present a new method that provides both protection of original texts and performing fast document similarity searches. Our system uses Burrows-Wheeler Transform (BWT) to create and store suffix arrays, and Multi-Level Alignment (MLA) to search similar regions with various user-defined precision settings. It also uses a lossy compression method so-called Korean Skin Extraction to protect the original sources. By experimental evaluation, we demonstrate that our proposed system effectively protects original texts as well as performs precise and fast searches with text queries longer than 20 letters. We also show that the proposed system can adjust trade-off between the inspection time and precision by user-defined settings with MLA.

주제어