摘要
日语口语语料库为用户提供约660小时自发言语的音频语料,以及转写文本、语素标签、语音标签等注释信息,具有规模庞大、语体多样、素材丰富、注释详尽等特点。在日本,该语料库广泛用于包括语法、语音在内的口语研究,已取得丰硕成果。在国内,虽然对日语口语进行了大量研究,但主要集中在对口语的文本分析,鲜有对口语的语音分析。基于该语料库丰富多样的信息,可以将语言内、外部要素联系起来,将言语行为与言语意识结合起来;可以从词性角度考察词形融合现象,从语言外部要素的角度考察语言内部结构的变化;可以从声学特征角度考察语音表现中的差异。从而扩充研究范畴、拓宽研究视角。
Corpus of Spontaneous J apanese provides users with about 660 hours of audio of spontaneous speech,as well as annotated information such as transcribed text,morpheme labels,and phonetic labels.It is characterized by large scale,diverse language styles,rich materials,and detailed annotations.In Japan,the corpus has been widely used in oral language research,including grammar and phonetics,and has achieved fruitful results.In China,although a lot of research has been done on spoken Japanese,it mainly focuses on the textual analysis of spoken Japanese,and rarely on the phonetic analysis of spoken Japanese.Based on the rich and diverse information in this corpus,we can connect the internal and external elements of language,and com bine speech acts with speech consciousness.We can examine the phenomenon of word-form fusion from the perspective of part of speech and the change of internal structure from the perspective of external elements of language.The differences in speech performance can be examined from the perspective of acoustic characteristics,thus expanding the research scope and broadening the research perspective.
作者
胡伟
Hu Wei(Dongbei University of Finance and Economics;Waseda University)
出处
《日语教育与日本学》
2022年第1期134-142,共9页
Japanese Language Education and Japanese Studies
基金
2020年度辽宁省社会科学规划基金项目“日语学习者历时语音语料库的建设及应用研究”(项目编号:L20BYY019)的阶段性成果
关键词
日语口语语料库
CSJ
自发言语
语音
Corpus of Spontaneous Japanese
CSJ
spontaneous speech
phonetic
