Development and Testing of Transcription Software for a Southern Min Spoken Corpus

Computational Linguistics and Chinese Language Processing Vol. 17, No. 1, March 2012, pp. 1-26 1 © The Association for Computational Linguistics and Chinese Language Processing Development and Testing of Transcription Software for a Southern Min Spoken Corpus Jia-Cing Ruan∗, Chiung-Wen Hsu∗, James Myers∗, and Jane S. Tsay∗ Abstract The usual challenges of transcribing spoken language are compounded for Southern Min (Taiwanese) because it lacks a generally accepted orthography. This study reports the development and testing of software tools for assisting such transcription. Three tools are compared, each representing a different type of interface with our corpus-based Southern Min lexicon (Tsay, 2007): our original Chinese character-based tool (Segmentor), the first version of a romanization-based lexicon entry tool called Adult-Corpus Romanization Input Program (ACRIP 1.0), and a revised version of ACRIP that accepts both character and romanization inputs and integrates them with sound files (ACRIP 2.0). In two experiments, naive native speakers of Southern Min were asked to transcribe passages from our corpus of adult spoken Southern Min (Tsay and Myers, in progress), using one or more of these tools. Experiment 1 showed no disadvantage for romanization-based compared with character-based transcription even for untrained transcribers. Experiment 2 showed significant advantages of the new mixed-system tool (ACRIP 2.0) over both Segmentor and ACRIP 1.0, in both speed and accuracy of transcription. Experiment 2 also showed that only minimal additional training brought dramatic improvements in both speed and accuracy. These results suggest that the transcription of non-Mandarin Sinitic languages benefits from flexible, integrated software tools. Keywords: Speech Transcription, Southern Min, Taiwanese, Romanization, Key-in Systems. ∗ Graduate Institute of Linguistics, National Chung Cheng University, Minshiung, Chiayi 62102, Taiwan Telephone: (05) 272-0411 ext. 21510; Fax: (05) 272-1654 E-mail: Lngmyers@ccu.edu.tw The author for correspondence is James Myers. 2 Jia-Cing Ruan et al. 1. Introduction 1.1 Constructing a Southern Min Speech Corpus As with any language, corpora of spoken Southern Min (Taiwanese) have many uses, both scientific and practical. Corpora of written Southern Min exist (e.g., Iunn, 2003a,b, 2005, based on novels, prose, dramas, and poems; the Southern Min Archives of Academia Sinica, 2002; Ministry of Education, 2010, with word frequency statistics), but Southern Min, unlike Mandarin, is virtually never written at all. For this reason, there has been increasing interest in corpora of spoken Southern Min, including the NCCU corpus of spoken Chinese (Chui, 2009), which includes everyday conversation in Southern Min, and ForSDat (Formosa Speech Database) of Lyu, Liang, & Chiang (2004), which is a multilingual speech corpus for Southern Min, Hakka and Mandarin. One area where a spoken corpus is essential is in the study of first language acquisition. This consideration motivated the construction of the Taiwanese Child Language Corpus (TAICORP; Tsay, 2007), which contains about two million morphemes in half a million utterances, based on about 330 hours of recordings of spontaneous conversations between children and their caretakers. Speech corpora are also essential for understanding the use of language in adult conversation, motivating our corpus of adult spoken Southern Min (Tsay & Myers, in …

Content truncated to 500 words