A Feasibility Research on Text Data Description Rule and Database for Japanese Classical Literature.

基于与SGML相同的概念，研究开发了一种适用于日本古典文学文本的文本数据描述规则Kokin(KOKubungaku Information)规则。利用该规则描述了许多日本古典文学文本，如岩上出版的《日本古典文学提纲》(100卷约600部作品)、东藤道《讲故事提纲》(20卷)和《日本诗集选集》21本。然后，利用新开发的数据模型，将这些数据组织、开发并作为全文数据库在日本文学研究所(NIJL)的主机上实现。其中部分数据还分别作为本地数据库服务器被组织和实施到工作站和PC机上，特别是用于故事大纲全文数据库的光盘是与工作站或PC机上的文本处理系统一起开发的。其中一个已经投入实际使用的系统被称为Media-Finder系统，从其起源和传播的情况来看，只将某一文献的一个文本作为数据准备的对象是极其困难的。为了在多个传送的版本之间相互参照，也许有必要准备包括大多数这些传送的版本的数据库。出于这个原因，其中排列和组合一些传输的版本的修正是可用的。此外，本文还对SGML在TDR上的可用性进行了评估。DTD(数据类型定义)是为与有关《唐卡选集》的SGML一起使用而创建的。然后，将数据存储到数据库中，并由SGML系统上提供的字符串搜索工具(例如Mark-It)使用。

A Text data Description Rule (TDR) called KOKIN (KOKubungaku INformation) rule has been researched and developed for Japanese classical literature texts, based on the same concept as the SGML.By this rule many Japanese literary texts have been described, such as the Outline of Japanese classical literature published by Iwanami (100 volumes.approx.600 works), the Outline of storytelling by Tokyou-do (20 volumes), and 21 Selections of the Anthology of Japanese Poems, Tanka.Then, these data are organized, developed and implemented as the full-text database using new developed data model to the main frame computer of the National Institute of Japanese Literature (NIJL). And, parts of these data are also organized and implemented to the Workstations and PCs as the local database server, respectively.Particulaly, CD-ROMs for full-text databases on the Outline of Storytelling are developed with text handling systems for the Workstations or PCs. One of these systems has been put to practical use, named as Media-Finder system.It is extremely difficult to take only one text of a certain literature as the subject of data preparation, from the circumstance of its origin and transmissions. For cross-reference among multiple transmitted versions, it is perhaps necessary to prepare databases that include most of these transmitted versions. For this reason, a recension, in which some transmitted versions are arranged and combined, is usable. So, the recension text databases are necessary for research in Japanese literature.Besides, availability on the TDR by SGML was evaluated. DTD (Data Type Definition) are created for use with SGML about the Anthology of Tanka. Then, the data were stored to database and used by a character string search tool provided on SGML system (ex.Mark-it).