ADAPTIVE DECENTRALIZED WEB SEARCH INSPIRED BY SMALL-WORLDS AND SWARM INTELLIGENCE

摘要

In this paper we propose a framework of a decentralized Web search system being organized as a peer-to-peer network. In such decentralized search system, the global manager or coordinator is absent; instead, the member search services connect with one another in a self-organizing mode, giving rise to the evolution of the entire search system or "search network". This architecture may mitigate the scalability problem the centralized search systems and meta-search systems suffer. However, connection management would become a key challenge. Following a metaphor of the human memory system, we envision to combine the "Small World" network model with an Ant Colony Systems algorithm to enhance connection management and query-routing of the decentralized Web search system.