BACKGROUND

Finding useful information on the web can be difficult. In 1997 this was becoming increasingly apparent, and more and more organizations began working to improve the situation by creating small, stand-alone directories of high-quality Internet resources. While these directories in many cases proved to be invaluable tools, in some sense they just pushed the problem up one level, requiring the user to find the pertinent directories and search each one for the desired resources. The Isaac project was born out of a desire to solve this problem.

PURPOSE

The Isaac Network provides a means to link these small, stand-alone Internet resource directories by allowing groups to link together geographically- or organizationally-disparate collections of Internet resource description metadata, providing users with the capability of searching through all of the linked collections via a single, coherent web interface. The base standard metadata fields for Isaac systems are based on Dublin Core, and allow information to be queried based an array of standard fields present in all collections, including title, subject, author, and description.

TECHNOLOGY

To insure compatibility and extensibility, the Isaac software runs on readily-available platforms and uses well-established protocols for data storage and intercommunication. The primary development and target operating system for Isaac is Linux. For directory queries and intercommunication Isaac uses LDAP, and for index generation and exchange, Isaac makes use of CIP.

THE FUTURE

In November 2000, after a ten-month lull, development has begun on version 2 of the Isaac software. Principle goals for the new version include:

Better Scalability - When deployed in its initial incarnation, the Isaac software requires a full-connected network, which limits long-term scalability. Version 2 of the Isaac software will include a hub / cache mechanism to remove this limitation.

Data Import Tools - The data that an organization may want to make available via the Isaac network may have been originally stored in many forms. Version 2 of Isaac will include a set of core tools intended to assist in converting data for use with Isaac.

Easier Installation - The goal is to make setting up an Isaac node a near-turnkey process, to minimize the technical resources required to make a collection of metadata available via Isaac.

The initial alpha release of Isaac v2 is expected to be available in the Spring of 2001. If your organization may be interested in participating in the alpha or beta testing or would like to make use of the production release of Isaac, please send a note to scout@cs.wisc.edu to let us know of your interest.

RESOURCES

A Distributed Architecture for Resource Discovery Using Metadata

Originally published in the June, 1998 issue of D-Lib magazine, this article gives a good overview of the technology and issues behind Isaac.

The Isaac Network: LDAP and Distributed Metadata for Resource Discovery

Presented at the 1999 IEEE Metadata conference, this paper covers the concepts and general principles used by Isaac, and briefly discusses some of the technological choices made during its implementation.

Isaac Demonstration Search Page

A search page set up for testing on top of a cluster of Isaac nodes running at systems hosted by Scout and several European collaborators.

CURRENT AND PROSPECTIVE ISAAC NETWORK PARTICIPANTS

MathGuide, Lower Saxony State and University Library G?ttingen SSG-FI

GeoGuide, Lower Saxony State and University Library G?ttingen SSG-FI

InSite , Cornell Law Library

Infomine, University of California

Curricular Resource Library, Project CuRL

LSU Digital Library, Louisiana State University

BUBL LINK, BUBL Information Service

History Matters, City University of New York /George Mason University

Scriptorium, Duke University Libraries

Internet Pointer Guide, Technical Knowledge Center & Library of Denmark (DTV)

Energy & Environmental Information Resources Center (EEIRC), University of Louisiana at Lafayette

Librarians' Index to the Internet, Berkeley SunSITE

EdNA, Education Network Australia

Edinburgh Engineering Virtual Library (EEVL), Heriot-Watt University Library

OMNI: Organising Medical Networked Information, University of Nottingham

SOSIG, UK Resource Discovery Network

AMICO, Art Museum Image Consortium

Internet Public Library, University of Michigan, School of Information

Marriage, Women and the Law, 1815-1914: Studies in Scarlet, Research Library Group

AgEcon Search: Research in Agricultural and Applied Economics, University of Minnesota Libraries

ELVIS!: Elec. Library Virtual Info. Services, Seneca College