Document search interface design for large-scale collections and intelligent access
Javed Mostafa · 2002
As the universe of documents has enlarged from those available via the online catalog to a larger cluster of databases and web-accessible resources, interfaces are being created that can search multiple document collections simultaneously. Also, searching for document surrogates is losing favor as more documents are digitized and distributed in full-text form. Availability of full-text makes it possible for document components such as tables, illustrations, citations, and references -- components that traditionally remained outside the scope of document searching -- to be considered and exploited by search interface designers. Additionally, due to the popularity of web-hyperlinking people are beginning to expect linking of documents across different collections based on common semantic or non-semantic attributes. Increased research activity on artificial intelligence techniques for document access is leading to more fundamental changes in document searching. It is now possible to delegate 100% of the search effort to online search agents. Agents have been also created for performing tasks such as selecting appropriate collections, refining queries, and sorting results to assist with searches conducted in distributed environments. The broad scope of the workshop is on the impact of above technological changes on search interface design.