A Multimedia Database for Colloquial English from Closed-captioned Videodiscs

本研究编制了一个英语口语多媒体数据库。多媒体数据库从存储在硬盘上的大量电影剧本中检索包含任意关键词的句子，然后在包含编码字幕（封闭字幕）的视频光盘上寻找相应的场景。目前多媒体数据库中记录的字数已超过100万字，超过了伦敦-隆德语料库的50万字。这个多媒体数据库的创新之处在于，它能够将正字法抄录与附有字幕的视频光盘上的相应场景联系起来。这些类型的光盘可以在电视显示器上同时显示字母、声音和图像。只要看一眼视频光盘上的场景，即使是非英语母语者也能很容易地理解演讲事件的背景。多媒体数据库是从事英语作为第二语言教学的人员的基本信息来源。它可以作为口语英语研究的基础，并提供一些对语言表现的见解。这种新型数据库将开辟以前不可能的新研究领域。下面的报告描述了制作多媒体数据库的方法，并评估了它对英语口语研究的有用性。

This study has compiled a multimedia database for colloquial English. The multimedia database retrieves sentences containing arbitrary keywords from a vast number of film scripts stored on a hard disk and then seeks the corresponding scenes on videodiscs which contain encoded subtitles (closed captions). Currently the total amount of words recorded in the multimedia database is more than 1,000,000 words, exceeding the London-Lund Corpus's 500,000 words.The innovative feature of this multimedia database is its ability to link orthographic transcriptions with the corresponding scenes on closed-captioned videodiscs. These types of videodiscs allow the simultaneous presentation of letters, sounds and images on a TV monitor. With a mere glance at a scene from a videodisc, the context of the speech event can be easily understood even by non-native speakers of English.The multimedia database constitutes a basic source of information for those engaged in teaching English as a second language. It serves as an infrastructure for studies in colloquial English and provides some insight into language performance. Such new types of database will open up new areas of research that have not previously been possible. The following report describes the methodology used to make the multimedia database and evaluates its usefulness for studies of colloquial English.