The Self-Managing Library

Software prevents scholarly schisms

IE Index Intel The establishment of an internal group dedicated to life sciences applications signals the company's awareness of a potentially important new market. PeopleSoft Inc. The release of PeopleSoft 8.8 should allow the company to build on the mostly positive response to PeopleSoft 8. Data Mining Now that they've caught wind of the U.S. Defense Advanced Research Projects Agency "total information awareness" strategy, critics are causing a public uproar that may undermine the government's plans. Once again, data mining becomes a lightning rod. INDEX LEGEND HOT — On the upswing NOT — Possible trouble ahead STATUS QUO — No change

The Massachusetts Institute of Technology (MIT) and Hewlett-Packard have implemented a new, Web-accessible system for storing, indexing, and disseminating the university's intellectual property. DSpace is an electronic, open source platform for storage and retrieval that lets MIT maintain its own virtual library of digitally rendered material.

All of MIT Press's output was recently compiled into an adjunct database that lets students and scholars search everything from out-of-print books to recent publications.

MIT officials have put their university at the forefront of a revolution to convert other campuses to similar digital libraries. According to Mackenzie Smith, associate director of MIT Libraries and the DSpace project director, MIT will make DSpace's software available to other campuses for free, as well as provide logistical support.

Networking the MIT system to other universities' repositories will make material more readily available to researchers and students regardless of their university affiliation. Talks of linking MIT with Cornell and Cambridge, for example, are already underway.

HP's market for such technology could be significant if the system proves successful in implementation. (IBM has enjoyed a well-developed digital library business for some time.) Even government agencies have been closely following DSpace's progress.

Here's how it works: With registered access, MIT faculty and students submit material to the database through an assigned Web portal based on the submitter's "community," or academic department. Submissions are then reviewed and, if accepted, archived into the digital library. The host institution is responsible for creating approval processes and committees, likely analogous to specified departments, research centers, or laboratories. Currently, DSpace can hold 2TB of data, but a plan to expand the system forecasts an eventual capacity of more than a petabyte (or approximately 1,000TB).

While host institutions must resolve nontechnical problems, such as copyright issues or their affected relationships with academic journals and societies, the DSpace design may finally give institutions what they've needed for too long: a tool to disseminate and organize credible, searchable information.

A thorough, technical analysis of DSpace at MIT is available at dspace.org.

— Jill Duffy

In this Issue: