Abstract
In this paper we present the SEWASIE system, a multi-level agent-based architec- ture for querying heterogeneous data sources integrated by means of ontologies. Main features of this system are: two level data integration scheme, a query tool that supports the user in formulating a precise query, integrated tools for negotiation and information monitoring, and an agent infrastructure that provides a unifying framework for the architecture. In this work we focus on the querying process, from the user interfacer to the query answering mechanism. We show how the use of ontologies integrates the user interface with the underlying agent architecture.