User Environment for Accessing Multilingual Digital Library

多语种信息环境是全球数字图书馆建设的重要内容之一。此外，在日本，通过互联网访问数字图书馆的信息还需要解决一些重要问题，例如字符变体和未包含在标准字符集中的字符。基于这些考虑，研究人员使用最初由图书馆和信息科学大学（ULIS）开发的称为MHTML的多语言HTML文档浏览技术解决了国际数字图书馆网络的信息访问支持环境。本研究与美国匹兹堡大学信息科学学院的研究人员合作，对多语种工具在全球信息网络环境下的可用性进行了研究，主要研究了以下几个问题：（1）通过使用MHTML技术，用户原则上可以在没有外来字体的WWW浏览器上显示任何外文文本。在本研究中，除了显示功能之外，还研究和开发了文本输入功能。(2)我们开发了一套用多种语言表达的日本古老民间故事。这本古老的故事集包括11种语言的11个故事。翻译工作不仅由ULIS的工作人员完成，也由网络上的志愿者完成。(3)一个日本汉字变体的词库被开发出来。该词库包括从两个主要的日本工业标准代码集中提取的变体集，总共包括12，000多个字符。(4)本文从多语言元数据描述和多语言元数据注册两个方面对多语言元数据进行了研究，这两个方面对于全球网络上的信息访问都是至关重要的，研究者们将在这方面继续努力。

Multilingual information environment is one of the most important aspects in the digital libraries for the global community. In addition, there are important aspects which has to be solved for information access in the digital libraries via the Internet in Japan, for examples character variants and characters not included in standard character sets. Based on these considerations, the researchers tackled information access support environment for international network of digital libraries using a multilingual HTML document browsing technology called MHTML that was originally developed in the University of Library and Information Science (ULIS). In this research, usability of the multilingual tool in the global information network environment was investigated collaborating with the researchers at School of Information Sciences, University of Pittsburgh in USA.In this research, the following issues were investigated :(1) By using the MHTML technology, users can display, in principle, any foreign texts on his/her WWW browser which has no foreign character fonts. In this research, text input functions were investigated and developed in addition to the display function.(2) We have developed a set of Japanese old folk tales expressed in multiple languages. This collection of oldtales includes 11 tales in 11 languages. Translation were done not only by staffs at ULIS but also volunteers on the network.(3) A thesaurus of Japanese Kanji Variants were developed. The thesaurus includes sets of variants extracted in two principal Japanese Industrial Standard code sets that include more than 12,000 characters in total.(4) We have investigated metadata in multiple languages in two aspects, which are metadata description in multiple languages and metadata registry in multiple languages.All of these aspects are crucial for information access on the global network and the researchers will make continuing effort on these issues.