Abstract:

Distributed information retrieval, also known as federated search, is an ad-hoc search problem in environments containing multiple, possibly many, on-line databases. The contents in those online databases are invisible to general commercial search engines like Google. Distributed IR is the tool to probe the invisible web contents. In these talk, we will briefly take about previous research in distributed IR task. Specifically, we will talk about some common resource selection algorithms and introduce some of our recent work in this field.

Related Readings:

Distributed Information Retrieval J. Callan In Advances in Information Retrieval, W.B. Croft editor, 2000 Using Sampled Data and Regression to Merge Search Engine Results L. Si, J. Callan In Proceedings of SIGIR 2002 Distributed Search over the Hidden Web: Hierachical Database Sampling and Selection P. Iperotis, L. Gravano In Proceedings of VLDB 2002