The transition from searching in a single database to searching a multiplicity of networked atabases exacerbates some old difficulties in the design and evaluation of retrieval systems and creates new ones. A network environment calls into question the traditional definitions of recall and relevance. Efficient network searching raises questions about where to look first, where to look next and when to stop searching. The need for “entry vocabulary” support and the need for support in moving from one system vocabulary to another are increased by the increased use of more different databases. The network environment offers the option of collecting different representations of the same object and merging them into an extended
[1]
Michael K. Buckland,et al.
Union Records and Dossiers: Extended Bibliographic Information Objects.
,
1994
.
[2]
Linda Schamber.
Relevance and Information Behavior.
,
1994
.
[3]
Irene P. Godden,et al.
Between Scylla and Charybdis: Subject Searching in the Online Catalog
,
1991
.
[4]
Christian Plaunt,et al.
OASIS: a front-end for prototyping catalog enhancements
,
1992
.
[5]
Luis Gravano,et al.
The Effectiveness of GlOSS for the Text Database Discovery Problem
,
1994,
SIGMOD Conference.