| 講演抄録/キーワード |
| 講演名 |
2025-11-04 10:10
[ショートペーパー]A Hybrid AI-Crowdsourcing Architecture for Large-Scale Digitization of At-Risk Cultural Heritage ○Chidchanok Choksuchat・Arafan Hemho・Sinchai Kamolphiwong(PSU)・Somrak Numnark(BIA)・Thichakorn Noo-Eid・Chanikan Saengkaew(PSU) IA2025-37 |
| 抄録 |
(和) |
This paper introduces the Dhammic Volunteerism Platform, a digital solution for preserving at-risk Thai Buddhist manuscripts. Our hybrid model synergizes advanced cloud OCR with community crowdsourcing from over 1,400 volunteers (as of Sep 17, 2025) to digitize 200,000+ pages of Buddhadasa Bhikkhu’s writings. A three-tier quality protocol, including AI transcription, peer review, and expert verification, ensures accuracy, while gamification sustains community engagement. Having transcribed over 14 million characters, this initiative has produced a first-unique and high-quality dataset. This creates a Data Flywheel for future innovations like APIs and specialized LLMs, offering a replicable blueprint for global heritage preservation. |
| (英) |
This paper introduces the Dhammic Volunteerism Platform, a digital solution for preserving at-risk Thai Buddhist manuscripts. Our hybrid model synergizes advanced cloud OCR with community crowdsourcing from over 1,400 volunteers (as of Sep 17, 2025) to digitize 200,000+ pages of Buddhadasa Bhikkhu’s writings. A three-tier quality protocol, including AI transcription, peer review, and expert verification, ensures accuracy, while gamification sustains community engagement. Having transcribed over 14 million characters, this initiative has produced a first-unique and high-quality dataset. This creates a Data Flywheel for future innovations like APIs and specialized LLMs, offering a replicable blueprint for global heritage preservation. |
| キーワード |
(和) |
Cultural Heritage / Digitization / Artificial Intelligence (AI) / Crowdsourcing / Optical Character Recognition (OCR) / Data Flywheel / / |
| (英) |
Cultural Heritage / Digitization / Artificial Intelligence (AI) / Crowdsourcing / Optical Character Recognition (OCR) / Data Flywheel / / |
| 文献情報 |
信学技報, vol. 125, no. 222, IA2025-37, pp. 9-14, 2025年11月. |
| 資料番号 |
IA2025-37 |
| 発行日 |
2025-10-28 (IA) |
| ISSN |
Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
| PDFダウンロード |
IA2025-37 |
|