面向计算机辅助正音的汉语中介语语音语料库的创制与标注

The Construction of a CAPL Chinese Interlanguage Corpus and Its Annotation

本文介绍面向计算机辅助正音的汉语中介语语音语料库的创制思路与标注方法。该库包括单音节、双音节、三音节、四音节、句子和语段六个子库。它最大的特点是在声母、韵母、声调及语调等方面进行了偏误标注。对声母、韵母,它将提供中介语发音人在发音部位和/或发音方法上的偏误信息;对声调、语调,它将提供声调特征点、调核、边界调、停延/间断等方面的偏误信息。这些信息将为汉语CAPL系统提供可靠、实用的正音线索。

This paper describes the design and annotation of a Chinese Interlanguage Corpus for Computer Assisted Pronunciation Learning steps differing in number of the samples. It is co （CAPL）. The corpus is created in two reposed of six sub-corpora： monosyllables, the one of disyllables, the one of tri-syllables, the one of quarter-syllabi one of es, the one of sentences and the one of paragraphs. Analysis of the errors in initials, finals, tones and intonations are presented by certain labels error information and also clues to making in the corpus. The data of the corpus can provide right pronunciation.