Text-image-speech Database in T.E.F.L.

本研究是对1991-1992财政年度同一资助项目的后续研究。计算机化的Cops在语言和文学研究方面都有很大的可能性。传统上，日语词典完全基于英日词典的视角，而英日词典则以个别词典编纂者收集的样本为基础。这导致了对英语单词和短语的误判和/或错误解释。计算机语言语料库可以纠正这些错误，通过使我们能够通过更多的定量评估来查看真正的英语(COBUILD词典的口号)，加深我们对文学文本的理解。本项目的目的是为个人研究人员提供一种方法，从书面和口头来源积累大量的真正的英语文本，并提供一个软件包，使得在这样的数据库中搜索关键词变得容易和快速。

The present research is a follow-up of the previous research project carried out during the fiscal years 1991-1992 with the fund provided by the same Aid-in-Grant.Computerized Copuses have enormous possibilities both in language and literature studies. Traditionally, T.E.F.L.have been entirely based on the perspective set out by English-Japanese dictionaries, which are in turn based on samples collected by individual lexicographers. This has led to misjudged and/or mistaken interpretations of English words and phrases.Computerized language corpuses can correct those errors and deepen our understanding on literary texts by enabling us to look at 'real English' (a slogan for COBUILD dictionaries) with more quantitative assessment.The present project has aimed at providing individual researchers with a way to accumulate huge amounts of 'real' English texts from both written and spoken sources, and with software package that makes it easy and speedy to search keywords in such databases.