Improving the Lookup Performance of Chord Network by Hashing Landmark Clusters

Yu, Yun-Shuai; Miao, Yu-Ben; Shieh, Ce-Kuen

doi:10.1109/icon.2006.302674

Cited by 6 publications

(1 citation statement)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In , the researchers proposed an associative search based on unstructured architecture, which retains property of semantic search and improves the efficiency to locate rare target resource. On the basis of structured P2P networks, eigenvector is introduced to map file properties in and , where vector space is constructed through operating on keywords, the distance between two vectors reflects node's interest similarity. Then, nodes are divided into multi‐level clusters, and the target files are located in the common interests cluster based on cluster index.…”

Section: Introductionmentioning

confidence: 99%

Double‐layer P2P networks supporting semantic search and keeping scalability

Liu

Wang

2013

Int J Communication

View full text Add to dashboard Cite

SUMMARYIn peer-to-peer (P2P) networks, with the aim to support semantic search and keep scalability at the same time, a novel double-layer P2P protocol is proposed. Firstly, keywords representing resource characteristics are compressed into bloom filters and structured P2P layer is constructed. Secondly, through analyzing one's bloom filters, each node extracts its interest, which is represented by a characteristic vector V . After obtaining common interest among nodes through comparing their V s, a small-world model is constructed, and another unstructured P2P layer is obtained. Consequently, on the basis of the decentralized architecture, a double-layer semantic search P2P protocol is structured. When objective resource is being located, its similarity to local interest is used to determine whether it should be located in unstructured P2P semantic layer or structured layer adaptively, which not only improves search efficiency but also supports semantic location more widely.

show abstract

Section: Introductionmentioning

confidence: 99%