A popular notion today is that of the Institutional Digital Repsoitory -- a space dedicated to cataloging, filing, and subsequently locating the intellectual output of the academy.
Enter DSpace -- an open source software product developed jointly by the Massachusetts Institute of Technology and Hewlett-Packard. DSpace is a groundbreaking digital library system designed to capture, store, index, preserve, and redistribute the intellectual output of a university in digital formats. DSpace is freely available to research institutions world-wide as an open source system that can be customized and extended. At UVM, DSpace is installed locally at http://badger.uvm.edu/dspace.
The organizational model of DSpace is one of Communities and Collections. Communities can be any arbitrary grouping of individuals or entities; e.g., Department of Geology, School of Natural Resources, Asian Studies Program, or Dr. Smith's Research Group.
Collections are groups of related material generated by a community: e,g., Plant and Soil Science might have collections related to entomology, soil science, horticulture, Greenhouse Educational Materials, and agronomy.
Finally, within a collection are items. An item consists of metadata and one or more bit streams. Metadata describes the item's attributes -- among them Title, Author, Language, format, abstract, keywords, publication date, etc. Bit streams can run the gamut from Word or PDF documents to PowerPoint presentations, still images, audio files, and videos. A single item might contain several bit streams: for instance, an item representing an historic postcard might contain a scanned image of the front, a scanned image of the back, and a PDF file containing a transcription of the handwritten message.
Some of the more compelling features of DSpace are the integrated search database and the concept of persistent Uniform Resource Idefinitiers (URIs).
The integrated search database allows all or part of DSpace to be searched for words appearing anywhere within an item's metadata. This -- along with traditional browse by author, title, or date capabilities -- provides powerful finding aids for locating anything or everything store in a DSpace repository. Additionally, DSpace can export an abstract of its contents to a variety of global search aggregators via a mechanism called Open Archives Initiative (OAI). This exposes our local UVM content to a wider audience.
Items in DSpace are assigned a simple and unique URI which remains constant for as long as the host institution decides to support DSpace, providing a permanent address on the web for an item. This facilitates "permanent" references to DSpace items from within bibliographic entries. For example, the URI of this image
http://hdl.handle.net/2051/2202
which currently points to:
http://badger.uvm.edu/dspace/handle/2051/2202
should remain unchanged for the life of the item, even if the UVM DSpace installation is moved to a different server (weasel.uvm.edu, for example).
DSpace is in use nationally serving content from MIT, Cornell, and many other institutions. Here at UVM, DSpace has been installed as a collaborative project between CIT and Academic Computing Services (now part of the new Learning Resources Group). UVM Collections of note now include the History HST 11/12 collection (images password protected), The Geology Slide Collection, and the Plant and Soil Sciences Greenhouse Education Collection. Says Buddy Tignor of Plant and Soil Science, "We have already used DSpace to archive several hundred greenhouse images with descriptions, over 3 hours of digitized video with associated transcripts, and software developed by greenhouse engineers specifically to aid students in understanding key greenhouse design concepts." DSpace forms the core of the World Greenhouse Education site, partially funded through USDA Grant 2003-03869.
Want more details, or to start your own collection? Contact the author.
Author: Wesley Wright, Academic Computing Services.
Last modified July 12 2005 09:46 PM
