Collaborative Classification of Large, Growing Collections with Evolving Facets A collaborative system is proposed to allow users to collaboratively organize a collection with an evolving faceted (multi-perspective) classification schema. Through users? collective manual efforts, and significantly assisted by the system, the faceted classification evolves with the growing collection, the growing user base, and evolving user interests. This proposed project will build a test bed using US Government Photo Multimedia Collection, a large federated multimedia collection containing some of the nation?s most precious multimedia documents. The current federation is practically inaccessible in terms of searching or browsing objects across multiple collections. The project will demonstrate that collaborative faceted classification is effective (e.g. better precision and recall) and efficient (e.g. less time to build and maintain) in supporting exploration of a large, growing collection of objects of various types. The final deliverables of this project include a system supporting collaborative faceted classification, tested on a significant sub-collection of the US Government Photo Multimedia collection. The system includes novel user interfaces, schema enrichment/construction algorithms and automated document classification techniques. The deliverables also include a tested methodology that allows instantiation of the same system to other large collections.

Based on this concept, we created a demo site. Please, visit the following link.

African American History Image Collection