AbIx: an approach to content-based approximate query processing in peer-to-peer data systems

摘要: In recent years there has been a significant interest in peer-to-peer (P2P) environments the community of data management. However, almost all work, so far, is focused on exact query processing current P2P systems. The autonomy peers also not considered enough. addition, system cost very high because information publishing method shared based each document instead set. this paper, abstract indices (AbIx) are presented to implement content-based approximate queries centralized, distributed and structured It can be used search as few possible but get many returns satisfying users' guarantee peers. Also, have low cost, improve speed, support frequent updates set method. order verify effectiveness indices, simulator 10,000 peers, over 3 million documents made, several metrics proposed. experimental results show that work well various