Active replication and update of content in electronic commerce

This paper describes a strategy for actively replicating and updating conte nt for electronic commerce. Active replication and updating of content is a chieved by intelligent agents (IA) using a time-invariant fragmentation app roach to partitioning and replicating data in a distributed computing envir onment. Taking into account the time-sensitivity property of data, IAs deri ve time-invariant fragments for their respective nodes from the query histo ry. A time-invariant fragment (TIF) is that portion of the database whose v alues do not change during a specified time interval. The algorithm that IA s use in creating TIFs for each node, for a given time interval, is present ed. The active replication approach is compared with three other approaches , Full-replication, nonreplication, and materialized view, in terms of data transmission costs. Results indicate that the active approach can be most effective for electronic commerce because of the high percentage of modific ation queries, the large sire of the network, and the greet number of trans actions.