The Delphi toolkit: Enabling Semantic Search for Museum Collections

Publication Type  Proceedings Article
Year of Publication  2008
Authors  Black, Michael; Schmitz, Patrick
Conference Name  Museums and the Web 2008. Proceedings
Conference Start Date  April 9-12, 2008
Publisher  Archives & Museum Informatics
Conference Location  Montreal, Quebec, Canada
Editor  Trant, Jennifer; Bearman, David
Keywords  semantic search; faceted browser; social media; community annotation; collections browser; open source
Abstract  

New technologies using Artificial Intelligence and Natural Language Processing are currently driving the booming enterprise search sector. These applications produce powerful, easy-to-use search and discovery tools. These 'semantic' tools and techniques have not been applied to museum collections. Some '2.0' have begun to see application in the museum domain, but have not been integrated into a common framework.

We describe the Delphi framework of semantic tools and community annotation for museum collections. The toolkit includes linguistic analysis tools, services that produce an easy-to-use faceted browsing user interface (UI) that makes it simple and fun to explore and understand museum collections. Personalization and social media tools allow creation and sharing of favorite sets of objects. The tools abstract the core technologies, and so can be used by designers and information architects without requiring specialized technical knowledge.

We deployed Delphi for the large collections at the Phoebe A. Hearst Museum of Anthropology with good success. We describe our experience with this deployment and lessons learned along the way. We explain the major components of the Delphi open source toolkit, and we discuss ongoing research that builds upon the platform.

URL  http://www.archimuse.com/mw2008/papers/schmitz/schmitz.html

Comments

Post new comment

The content of this field is kept private and will not be shown publicly.
CAPTCHA
This question is for testing whether you are a human visitor and to prevent automated spam submissions.
Syndicate content