A Uniform Interface for Active and Intelligent Information Access and Discovery

We propose an innovative combination of artificial intelligence (AI) techniques in the design of a software agent based tool for the discovery of information in networks of heterogeneous information resources. Specifically, our agent would act as a meta level interface to database resources in order to aid users in data analysis tasks. Within this framework we will offer tools which enable interactive "data archeology" as well as fully autonomous information discovery. This tool will allow users to manipulate large amounts of data and uncover implicit patterns with a minimum of effort. By utilizing multiple knowledge discovery algorithms in the execution of a task, and integrating the results, our approach presents a scaleable method for reliable information discovery. We further propose to provide Active Data Mining techniques that can monitor data and alert users to growing trends. Additionally, we will offer a unique dynamic interface which allows users to efficiently specify and reconfigure search tasks. Phase I research and development of a proof of concept limited prototype will lay the groundwork for the phase II implementation of a fully functional prototype of an intelligent knowledge discovery agent.