2026-10-12T18:28:00Z https://www.scidb.cn/oai

10.57760/sciencedb.28659 2026-07-13T13:53:48Z

2026-07-13 A Sentence-Level Corpus of Naxi Dongba script doi:10.57760/sciencedb.28659 en The Dongba Sentence Translation Corpus is used for machine translation of Dongba sentences and is stored in a structured folder hierarchy.This corpus is stored in a structured folder hierarchy. The corpus contains 1,148 sentence folders and a total of 2,138 characters/words of Dongba script. Each sentence folder follows the naming format “con<number>” (e.g., folder “con0001”) and includes the following files:(1) Dongba script character image: Stored in PNG format, named as “<number>.png” (e.g., “01.png”).(2) DB code sequence file: Stored in TXT format and named “word<number>.txt” (e.g., “word0001.txt”), with DB codes separated by spaces.(3) Chinese sentence translation file: Stored in TXT format and named “chinese<number>.txt”, (e.g., “chinese0001.txt”).For example, the folder for sentence “con0001” contains the DB code sequence “DB426 DB528 DB58 DB1351 DB1491 DB127 DB169”, the Chinese sentence translation “在郭上有两颗白石。”, and individual definitions for each DB code.  corpus; Dongba; sentence translation dai wei jia Li Xiaoliang Deng Fei Chen Shanxiong PUBLIC https://creativecommons.org/licenses/by/4.0/ dataset Science Data Bank