Study on New Digital Library Services Using Document Images and the Web

点击翻译按钮获取中文摘要

Digital libraries which store academic papers require some techniques for automatic extraction of bibliographic and other information from the papers. This study proposed a method of automatically extracting bibliographic information from OCRed document images of title pages of academic papers and from reference strings listed at the end of papers by using a conditional random field (CRF). This study also proposed a recommendation service of related papers using the technical terms extracted from academic papers and investigated ways to support online-browsing of academic papers using tablet PCs.