Research on Knowledge Extraction from Ancient Mongolian Historical Documents using Deep Learning

2020 Fiscal Year Final Research Report

• Project/Area Number: 17K00457
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Library and information science/Humanistic social informatics
• Research Institution: Ritsumeikan University
• Principal Investigator: BATJARGAL BILIGSAIKHAN 立命館大学, 衣笠総合研究機構, 研究員 (30725396)
• Project Period (FY): 2017-04-01 – 2021-03-31
• Keywords: deep learning / historical documents / traditional Mongolian / machine learning

Outline of Final Research Achievements

In this research, an information extraction and analysis method for digitized ancient Mongolian historical documents is proposed. The proposed method extracts features from historical manuscripts by utilizing deep learning techniques. Extracted deep features are utilized for building retrieval systems that encode the interpretations of ancient Mongolian words, as well as features for recognizing different letters with the same shape. Digital representations of ancient historical manuscripts with annotated ancient Mongolian texts along with the scanned images of manuscripts could be used as scholarly tools. The proposed methods were applied not only to Mongolian historical manuscripts but also to Oracle bone script and Ukiyo-e, a Japanese traditional art.

Free Research Field

Digital Humanities

Academic Significance and Societal Importance of the Research Achievements

本研究は歴史的文書の分析にかかる時間と手間を軽減することができると考えられる。現代モンゴル語の文書には含まれない隠れた知識を伝統的モンゴル文字の古文書から発見できると考えられる。提案方法を用いた歴史的文書の分析結果を分かりやすく表示するシステムはどんなユーザーにも利益をもたらすことが期待されている。さらに、歴史書類のデジタル化を研究対象にしている学者に大きく貢献すると考えられる。