japanese

Resource Sharing System

Resource Sharing System's nihuINT(Integrated Retrieval System)

1. System Configuration: Basic Principles and Elements

The Resource Sharing System is made up of two systems: the nihuINT(Integrated Retrieval System) and the nihuONE System. These systems were developed for the first term (FY 2005 – FY2007) and have been available to the public since April 2008.

Here, an overview of the Resource Sharing System will be given using the nihuINT(Integrated Retrieval System) as the primary example. For additional details, please refer to the following references:

*The Symmposium Lecture Material dated March 14, 2008pdfjapanese

2. Overview of the Entire System

The following eight principles were established for designing the system.

2.1. The Eight Principles

1. Target the original databases of the six institutes that have been made available to the public for sharing (108 as of April 2008).

2. Do not alter the original databases. In other words, do not reconstruct them or the like.

3. Although users will primarily be researchers in the humanities, make the databases available to the public.

4. Conduct a unified search - in other words, a cross search - of all the databases. Do not recognize the differences in the operating methods or the locations of individual databases.

5. Apply advanced processing to searches and results (using space-time data analysis systems and the like).

6. As much as possible, use technology that meets international standards.

7. Put the results of earlier research and the like to practical use.

8. Although the six institutes* will be targeted for now, consider connecting with other institutes in future.

 

Based on these principles, the nihuINT(Integrated Retrieval System) was constructed primarily with the following three technology components.

2.2. Three Technology Components

1. Use XML for standard data description.

2. Absorb the differences between databases using metadata (Dublin Core, primarily).

3. Unify search procedures using search protocols of an international standard (Z39.50 and SWR: Search/Retrieve Web Service).

3. Overview of the Entire System

Figure 1 provides an overview of system configuration. It is the overall concept of the initial plan. Each institute has already constructed its own databases, and service is provided by means of a dedicated information system. This information system is preceded by a front-end system (FES) and a gateway system (GWS).

The FES uses metadata to absorb the differences in the original databases. The GWS is in charge of user searches. Usually, users will conduct their information searches from the Web, through the GWS.
By coordinating the FES and GWS, cross searches are done with consistent standards. With the current system, there is no connection with external research institutes outside this organization, but this will be handled by consolidating metadata by means of harvesting. Overviews of the individual systems follow.

Overview of system configuration in the initial plan

Fig. 1 Overview of system configuration in the initial planjapanese

4. Individual Systems

(1) Basic System

Figure 2 shows the configuration of the current system. What is called the basic system is the combination of each institute’s FES and GWS.  The FES matches metadata for each original database and constructs and retains of individual metadatabases (MDBs). The FES connects to the GWS through information search protocol Z39.50 and the Search/Retrieve Web Service (SRW) and searches the MDBs for information. It works through Internet communication in which the GWS is the client and the FES is the server. Through Web access, users request cross searches on a Web server with GWS and obtain shared search results. The Web server provides the standard search system client and gateway functions and conducts the searches.  Z39.50 (Version 2) and SRW (Version 1.1) explicitly name each database but are basically standardized with the SRW. The nihuINT(Integrated Retrieval System) has been developed using Infocom Corporation’s product, InfoLib-GlobalFinder.

Overview of basic system configuration at present

Fig. 2 Overview of basic system configuration at presentjapanese

(2) MGR

MGR (Manager) is one of the basic systems used at the organization’s headquarters. It handles operations management for the entire nihuINT(Integrated Retrieval System). It grasps the status of registration, deletion, updating and the like for MDBs; its operations management includes usage statistics, bottleneck measures, resource management, backup and the like. It also has functions for positioning common shared information resources (such as maps, place-name dictionary data and other such content, as well as software with higher-level processing functions, etc.) as servers and functions to accommodate user downloads. It has repository functions, as well. MGR is based on the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH). At present, there is no connection with external institutions, but it is used for acquiring operations management information from institutes’ FES.

(3) Metadata

Based on Dublin Core, metadata formulates mapping rules3 for National Institutes for the Humanities. It creates MDBs for the over 100 databases of the six institutes. It searches and displays in six categories: all items, 5W1H (who, what, when and where), bibliographies, (names, authors and titles), details (5W1H + bibliographies), and extended DCMES (Dublin Core Metadata Element Set). Additionally, it defines space-time metadata for the GTA (Geo-Temporal Analyzer, a space-time data analysis system).

(4) User Interface

With consideration given to general use, a user interface was implemented that gives priority to ease of use while accomplishing demonstration experiments, monitor testing and the like. It handles search methods in the aforementioned five categories as well as more advanced searches geared toward specialists. Numerous functions related to search results are also provided, including list display, detailed display, result notes, keyword-highlighted display, and links to original databases.

Search screen progression

Fig. 3 Search screen progressionjapanese

Figure 3 shows the shifts in the search and display screens. Figures 4 and 5  show actual examples of search screens.  Links to the original databases are usually made by referring to URLs. The GWS has been equipped with resolver functions to act as relay functions to create and refer to URLs for linking when URLs cannot be created owing to FES metadata conversion or when the URLs to be referred to are made complicated by full data output or the like. An interface for administrators is also provided. Operations management of the basic systems environment and total operations management of the nihuINT(Integrated Retrieval System) through MGR are handled.

(5) GTA

A GTA (Geo-Temporal Analyzer) interface has been provided, in the form of a Web plug-in, not only for cross searching but as a higher-level process for searches and results. GTA search functions can be embedded into search screens, selected and used.  Chrono-oriented interface function (CIF) and geo-oriented interface function (GIF) are provided. Databases can be searched by specifying a time range using time slider or specifying a position range using a map. In addition, search results can be displayed on chronologies or maps. For additional details on GTA functions, please see the reference materials.

Example of a search screen

Fig. 4 Example of a search screen

Example of a screen listing search results

Fig. 5 Example of a screen listing search results