As a professor and former director of historical geography, and a director of university library in the past two years, I came to understand the importance and significance of the combination of research and practice, producers and consumers, authors and readers. The trend of digitalization and computerizations promote sharp and radical changes for all the libraries, especially those in universities. It also provides wide area and more opportunities for all the researchers and producers, include ECAI.

The Resource Sharing Information System and Metadata Suites: As the result of research activities (e.g., field works, literature searches, observations, etc.), variety kinds of databases have been created. These databases are expected to be the basis of Area Studies, but most of them are rather disappointing. As, each database uses its specific data structure and retrieval operations, then users must understand specifications of all databases (e.g., data structures, operations, locations, etc.) they want to retrieve data. The Resource Sharing Information System (RSIS) is the framework to integrate many databases on the Web seamlessly, and it enables users to retrieve many databases by one operation without concerning each database specification. One of the key technologies of RSIS is standard metadata. The first version RSIS uses Dublin Core Metadata (DCM). However, DCM defines only 15 data elements and is insufficient to describe complex spatiotemporal attributes of data, then “Coverage” element is expanded proprietarily to have some sub-elements for spatiotemporal attributes. The latest version of RSIS introduces Metadata-Suite which is the series of metadata from specific to general purposes. At present, specific metadata for each database, EAD, MODS, METS and DCM construct a basic Metadata-Suite.

Spatiotemporal Tools: Two spatiotemporal tools have been developed to analyze data. The first tool is HuMap (Humanities Map). HuMap is derivative software of TimeMap, but the latest version has many GIS analytical functions, and source codes are completely original. There are two types of HuMap. The one is a server application. This is the GUI for RSIS which runs on Web viewers to retrieve data by location and time. The other is the client tool which is a GIS tool and runs on PCs. The client tool can overlay many maps and images according to their locations, retrieve features from each layer, execute logical operations on features among layers, and process basic GIS analytical functions such as buffering and clipping. Like TimeMap, the client tool has an ability to process time attributes. Some characteristics of the client tool are database function and Java-plug-in function. The database function is used to retrieve and import data from H-GIS Data Clearinghouse. The Java–plug-in function links users’ Java programs and HuMap layer data, which will be the framework of HuMap to expand its GIS analytical functions.

The second tool is HuTime. It is revolutionary time-oriented software. As HuMap treats maps as layers, HuTime treats historical tables as layers. HuTime can overlay many layers according their timelines, retrieve events from each layer, and execute logical operations among layers.

Due to the different development phases, metadata used by both tools are different from each other, and then present HuMap and HuTime cannot use same spatiotemporal data simultaneously. Mutual metadata for HuMap, HuTime and H-GIS Data Clearing House is under construction, which will be used to operate HuMap and HuTime seamlessly and connect both applications with H-GIS Data Clearing House.

Ontology Database: Two ontology databases have been created to support spatiotemporal data organization. The one is The Gazetteer Database on Japanese Classical Places. It is the database of place names (rivers, lakes, mountains, shrines, temples, houses, monuments, villages, towns, counties, states, etc.) of present and past with their attributes and locations (longitudes and latitudes). The other ontology database is the calendar table. All dates described by different calendars are grouped and ordered according to Julian dates. This simple table can be used to convert a date of a calendar to the date of another calendar. We will open metadata definitions of these databases to expand data contents by cooperation with other institutes.