Emending and databasing of historical earthquake documents in the Japanese ancient and medieval ages and development of an open-access system

自公元6世纪左右以来，日本已知的几乎所有的历史地震记录都是以29卷的印刷书籍出版的，这些记录来自各种历史来源的剪报。它们是日本历史地震学和火山学的宝贵原始资料，但未知的史料仍需进一步搜索和补充。然而，这些书中有两个严重的问题。其一，它们包含了相当多不可靠的材料，因为编撰者没有对历史来源进行任何批判性的审查，也没有进行任何选择，这往往会导致历史地震目录的错误。为了从根本上改善这种涉及古代和中世纪的情况，我们通过地球科学家、历史学家和信息科学家的跨学科合作，批判性地审查了现有藏书中所有相关的历史文献，对每一篇文本进行了仔细的校订，并以数字形式(通过XML)组成了重新编辑的集合。然后，我们建立了一个古代和中世纪(直到公元1600年)所有历史地震/火山文献的全文数字数据库，并开发了一个开放获取系统，不仅供作曲家，也供最终用户将来改进数据库。虽然我们将研究对象限制在古代和中世纪，因为这些时代的史料与近代的史料有一些不同的特点，而后期的史料数量太大，不是一个小研究小组可以处理的，但我们项目中建立的方法和技术可以有效地应用到近代早期。该数据库即将通过互联网发布。

Almost all of Japan's known historical earthquake records since around the 6th century, which are clippings from various historical sources, have been published in type-printed books of 29 volumes. They are invaluable primary data for the Japanese historical seismology and volcanology, although unknown historical materials should further be searched for and added. There are, however, two serious problems in these books. One is that they contain rather many unreliable materials because the compilers did not carry out any critical examination of historical sources nor selection, which tends to produce errors in historical earthquake catalogs. The other is that the books are just a huge amount of printed papers, which makes full-scale utilization of massive information including keyword searching impossible.In order to improve these situations fundamentally concerning the ancient and medieval ages, we critically examined all the related historical documents in the existing collections, selected them, carefully emended every text, and composed a re-edited collection in digital form (by XML) through interdisciplinary collaboration among earth scientists, historians and information scientists. Then, we constructed a full-text digital database of all historical earthquake/volcano documents in the ancient and medieval ages (up to ca. A.D. 1600) and developed an open-access system for future improvement of the database by not only the composers but also end-users. Although we restricted the target within the ancient and medieval ages because historical sources in these times have somewhat different characteristics from those in the early modern age and historical records in the latter age are too huge amount to be treated by a small research group, the methodology and technique established in our project can be applied effectively to the early modern age. The database is to be released through internet soon.