臺灣客語語料之數位化

Abstract

本文旨在闡述臺灣客語語料庫之語料數位化，耙梳其流程整體脈絡並廓清文本授權與客語用字問題。語料作業流程係由「前置作業」與「數位化及檔案管理」兩大階段串聯，在「前置作業」中包含「語料盤點」、「語料徵集與授權」兩大步驟；而「數位化及檔案管理」則囊括「語料建檔與後設資料標註」、「語料數位化與資料清理」（含語料轉寫校訂）和「語料儲存與管理」三個部分。臺灣客語語料庫的重要性在於其為臺灣第一個書面語料與口語語料兼具且附口語錄音檔的帶標記語料庫，以系統化方式收錄臺灣客語六腔語料。藉由臺灣客語語料庫實際建構經驗，本文希望能發揮「鑒往知來」之效，提供其他專家學者參考，以應用到臺灣其他語言之語料庫建構，更希冀能為語言學與資訊科學之跨領域研究開創新機。

This paper lays out the digitization of corpus data in Taiwan Hakka Corpus, resolving the issues of texts authorization and Hakka character at the same time. The main task encompasses two stages: “preprocessing operation” and “digitization of corpus data and document management”. Taiwan Hakka Corpus with both written and spoken varieties (audio recordings available) of Taiwan Hakka language collected in a systematic manner is the first part-of-speech-tagged corpus among Taiwanese native languages. Its construction has taken the initiative in setting a model for corpus construction of other national languages in Taiwan. This paper demonstrates a significant reference for the development of interdisciplinary research on linguistics and computer science.