Dimension and Fact Mappings
in Peer Data Warehouses
Abstract: A peer-to-peer (P2P) data management system is made up of network nodes, called peers, which contain autonomous data sources and exchange data with other peers, called acquaintances. Peers are fully and independently responsible for the maintenance of their own data and the selection of acquaintances for data exchange. We assume that there is no global schema. Moreover, we assume that there is no centralized data source. A peer data warehouse (PDW) is a traditional data warehouse which is associated with a peer. Therefore, if a query issued at a local peer requires data from its acquaintances, the local peer must have appropriate mechanisms to retrieve the required data. In order to appropriately answer queries in a network of peer data warehouses, we must map dimensions and facts of acquainted peer data warehouses. We present a distributed algorithm for establishing those mappings.