Sciweavers

Share
WISE
2009
Springer

Aggregation of Document Frequencies in Unstructured P2P Networks

11 years 9 months ago
Aggregation of Document Frequencies in Unstructured P2P Networks
Peer-to-peer (P2P) systems have been recently proposed for providing search and information retrieval facilities over distributed data sources, including web data. Terms and their document frequencies are the main building blocks of retrieval and as such need to be computed, aggregated, and distributed throughout the system. This is a tedious task, as the local view of each peer may not reflect the global document collection, due to skewed document distributions. Moreover, central assembly of the total information is not feasible, due to the prohibitive cost of storage and maintenance, and also because of issues related to digital rights management. In this paper, we propose an efficient approach for aggregating the document frequencies of carefully selected terms based on a hierarchical overlay network. To this end, we examine unsupervised feature selection techniques at the individual peer level, in order to identify only a limited set of the most important terms for aggregation. We...
Robert Neumayer, Christos Doulkeridis, Kjetil N&os
Added 08 Mar 2010
Updated 08 Mar 2010
Type Conference
Year 2009
Where WISE
Authors Robert Neumayer, Christos Doulkeridis, Kjetil Nørvåg
Comments (0)
books