Research on Retrieval Tools with Ajax Technology on the Database of Old Korean Books existing in Japan

本文的研究目标是利用Ajax技术开发古代文献图像数据库系统和检索工具。为了添加这些文件图像和元数据，我们必须重新构建日本现存的旧韩文图书数据库DOKB。该图像库必须与古代文献刻录者名称库KOKUSEIU和作者印章阴影图像库IN-EI集成到当前的DOKB中。在这两年中，我们进行了以下的程序并得到了接下来的结果。第一年我们输入了4561张古代文献照片图像和11135张古代文献榆树图像。我们了解到，在古代文献的照片图像中，由于许多印章阴影图像与同一文献中的汉字字符重叠，因此很难从照片图像中剪切出作者的印章阴影。因此，我们很快就实现了包含古代文献刻字者姓名信息的数据库系统KOKUSHU。我们使用Ajax技术实现了Unicode汉字字符输入支持工具，并将其安装到名为DOKB.3的现有数据库系统中。我们开发了试用版输入支持工具，每次输入一个字符猜测一个关键词，并在下拉菜单区显示选定的候选关键词。古代文献图像数据库系统已经完成了大约70%的开发。在解决了如何从同一文档图像中截取印章阴影图像后，作者印章阴影数据库系统的开发停止了。目前，我们正在通过有效的技术和网络工具继续改进现有的DOKB数据库系统。

The goal of this research is to develop the ancient document image database system and retrieval tools with Ajax technology. We have to restructure the database of Old Korean Books existing in Japan named DOKB for adding these document images and its metadata. That image database must be integrated into the current DOKB with ancient document gravers name Database named KOKUSEIU and authors' seal shadow image database named IN-EI.We carried out the fallowing procedure and got the next results in these two years.1. First year we input 4,561 sheets image from ancient document photographs and 11,135 sheets image from Elms of ancient documents. We understood that it is difficult to cut out an author seal shadows from photo image of ancient documents because of the many seal shadow images were overlapped with Kanji character in same document. So, fast of all we have implemented the database system named KOKUSHU which is including ancient document gravers' name information.2. We have implemented the Unicode Kanji character Input Support tools with Ajax technology and installed into existing database system named DOKB.3. We have developed the trial version input support tools to guess a key word by input each one character every time and selected candidate key words are indicated at pull-down menu area.4. The ancient document image database system have been finished about seventy percent of the development.5. The development of the database system of authors' seal shadow image was stopped after solving how to cutout a seal shadow image from the same document images.At present we are continuing to improve this existing DOKB database system for better database system by effective technologies and web tools.