System components for embedded information retrieval from multiple disparate information sources

Abstract

Current information retrieval interfaces only address a small pant of the reality of rich interactions amongst user, task, and information sources. We view information gathering as an interactive, iterative activity involving multiple disparate information sources and embedded in the context of broader processes of information use. We have developed two key system components that enable information workspaces that adhere to this reformulation of information retrieval. The tlrst is a design for a user/system interaction model for retrieval from multiple, disparate information sources. The second is a repository modeling system, called Repo, that represents meta-information about different information repositories in a manner that supports system operation as well as provides a direct information resource to the user. To test these embodied the interaction system called Labrador.