A Self-organizing System for Large-scale Content-based Information Retrieval
| Authors | |
|---|---|
| Year of publication | 2008 |
| Type | R&D Presentation |
| MU Faculty or unit | |
| Citation | |
| Description | We propose a self-organizing system for content-based information retrieval which operates in an ordinary peer-to-peer network. The system is universal and allows us to search for various data types, e.g. multimedia, because we use the metric space data model. The self-organization of the network is obtained by using the social-network paradigm. The connections among peers in the network are created as social-network relationships formed on the basis of a query-and-answer principle. The knowledge of answers to previous queries is exploited to fast navigate to peers, possibly containing the most relevant answers to new queries. At the same time, a randomized mechanism is used to explore new and unvisited parts of the network to provide sufficient information for future exploitation. The proposed concepts are verified using a network consisting of 2,000 peers containing descriptive features of 10 million images from CoPhIR collection. |
| Related projects: |