Parallel corpus creation system and parallel corpus creation method

摘要

PPROBLEM TO BE SOLVED: To provide a system capable of efficiently collecting a large amount of bilingual text data used for a statistic machine translation system. PSOLUTION: In a bilingual corpus creation system, a translation target database 4 stores text data to be translated, and a server 1 publishes a webpage that displays the text data extracted from the translation target database 4 and an input box for receiving an input of a translation sentence to the text data on the Internet. The server 1 reads the translation sentence written in the input box from a terminal connected to the Internet for the published webpage to store the translation sentence in a translation sentence database 5, determines the correctness of the translation sentence when a fixed amount of the translation sentence is stored, and stores the translation sentence determined to be correct as text data of a bilingual corpus. PCOPYRIGHT: (C)2010,JPO&INPIT