Class-based approach in semantic P2P information retrieval

Peer-to-Peer (P2P) approach in information retrieval systems has drawn significant attention recently. P2P networks provide obvious advantages like scalability, reliability and, therefore, recent researchers are looking for a way to adapt these techniques to Information Retrieval fashion of nodes with heterogeneous documents. The greatest attention is paid to different semantic-based searching such as Gnutella Efficient Search (GES) proposed by Zhu Y et al., which derives from Vector Space Model. This paper proposes conceptual design of P2P unstructured information retrieval (IR) with heterogeneous documents on independent nodes.