We
propose a mix of basic and applied research to understand issues and create
compelling use-case demonstrations at the intersection of:
-
institutional
information management and digital asset management
-
personal and
collaborative information management
-
the
semantic web
We
seek to enhance interoperability among asset, schemas, metadata, and services:
-
across
distributed individual, community, and institutional stores, and
-
across
value chains that provide useful services by drawing upon the assets, schemas,
and metadata held in such stores
The
work hinges upon collaboration among:
-
MIT Lab for
Computer Science
(David Karger, principal
investigator)
-
The
Worldwide Web Consortium
(Tim Berners-Lee, co-PI; Eric
Miller, lead researcher)
-
MIT
Libraries
(Ann Wolpert, co-PI; MacKenzie
Smith, lead researcher)
-
Hewlett-Packard
Company
(HP Labs Bristol)
The
research methodology will include establishing a research sandbox to extend the
open-source DSpace system currently in Beta at MIT Libraries. We will:
-
extend
DSpace to greatly enhance its ability to support arbitrary schemas, and
metadata, primarily through the use of RDF and semantic web techniques
-
explore
support for distributed stores of assets, schemas, and metadata
-
implement a
dissemination architecture based upon web standards, enabling services to
operate upon relevant assets, schemas, and metadata within the distributed
stores
-
leverage
the haystack schemas and user interface to manage human interactions with the
personal, community, and institutional stores
To
ensure compelling results and impact we will:
-
establish a
critical-mass, co-located team in Cambridge, MA
-
ground the
research by focusing upon well-defined, real-world use cases in an exemplar
domain: research libraries
-
selectively
migrate demonstrated research results to DSpace operations
Since
parallel work is underway to deploy DSpace at a number of leading research
libraries, such an approach may lead to a powerful deployment channel through which
emerging standards can be efficiently applied by a supportive community.
The
research will be funded by invent@mit – the HP/MIT Alliance – and will be
conducted over a period of three years.
Alliance funding will total four million dollars.