Toward large scale integration: Building a MetaQuerier over databases on the Web

The Web has been rapidly "deepened" by myriad searchable databases online, where data are hidden behind query interfaces. Toward large scale integration over this "deep Web," we have been building the MetaQuerier system- for both exploring (to find) and integrating (to query) databases on the Web. As an interim report, first, this paper proposes our goal of the MetaQuerier for Web-scale integration- With its dynamic and ad-hoc nature, such large scale integration mandates both dynamic source discovery and on-thefly query translation. Second, we present the system architecture and underlying technology of key subsystems in our ongoing implementation. Third, we discuss "lessons" learned to date, focusing on our efforts in system integration, for putting individual subsystems to function together. On one hand, we observe that, across subsystems, the system integration of an integration system is itself non-trivial- which presents both challenges and opportunities beyond subsystems in isolation. On the other hand, we also observe that, across subsystems, there emerge unified insights of "holistic integration"- which leverage large scale itself as a unique opportunity for information integration.