Creating Spatiotemporal Rubbing-database for Discovering Potential knowledge

2020 Fiscal Year Final Research Report

• Project/Area Number: 18K18337
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 90020:Library and information science, humanistic and social informatics-related
• Research Institution: Ritsumeikan University
• Principal Investigator: meng lin 立命館大学, 理工学部, 准教授 (60615938)
• Project Period (FY): 2018-04-01 – 2021-03-31
• Keywords: 深層学習 / 古代文献整理 / 時空間データベース / 可視化 / 潜在的知識の発見

Outline of Final Research Achievements

This project aims to create a spatiotemporal rubbing database and apply keyword search for realizing the keyword visualization. The project may help for discovering potential knowledge such as historical politics, economics, and culture. The contributions are shown as follows.
(1) A deep learning and big-data analysis combined approach has been proposed for realizing the high accuracy rubbing character recognition. The proposed approach also supplementing the shortcomings of deep learning.(2) A spatiotemporal rubbing database has been created, which may help to re-organize the ancient documents.
(3) The visualization of keywords is realized, the results may be displayed the changes of keywords in spatial and temporal. (4) All of the functions are equipped on the server. And the API is realized, can be used freely for all the users.

Free Research Field

画像認識

Academic Significance and Societal Importance of the Research Achievements

研究成果の学術的意義としては、(1) 深層学習の不足を補う技術を提供し、高精度な拓本文字の認識を実現した。(2) 拓本と古代文献解析において、時空間データベースと可視化の機能を提供している。それにより、歴史に関する政治、災害、文化などの研究に対して、研究時間の短縮と効率化などに貢献できる。
社会的な意義は、時空間拓本データベースで検索と可視化を実現したことにより、古代文献の整理と潜在的な知識の抽出に貢献できたと考える。