Locality preserving scheme of text databases representative in distributed information retrieval systems

Abstract

This paper proposes an efficient and effective "Locality Preserving Mapping" scheme that allows text databases representatives to be mapped onto a global information retrieval system such as Peer-to-Peer Information Retrieval Systems (P2PIR). The proposed approach depends on using Locality Sensitive Hash functions (LSH), and approximate min-wise independent permutations to achieve such task. Experimental evaluation over real data, along with comparison between different proposed schemes (with different parameters) will be presented in order to show the performance advantages of such schemes. © 2010 Springer-Verlag Berlin Heidelberg.