Mining Very Large Databases Using Software Agents

Some databases are simply large (e.g., with ter_abytes of data). A number of potential and diverse patterns in databases would sufficiently be mined so as to support various decisions in applications. This research advocates a re-recognition and re-cogitation to how to discover very large databases. It also presents a new technique mining very large databases will be developed on different hierarchy for possible different applications. To do so, a system MVLDB (MiilingVery Large Databases) for such as analyzing, partitioning, re-organizing, and mining databases is <bulle by classifying and clustering in this paper. The system uses software agents in all aspects including sampling, data clustering and data mining.