Monitoring patterns through an integrated management and mining tool

Abstract

Patterns upon the data of many real applications are affected by changes in these data. We employ Pattern-Miner tool to detect changes of clusterings extracted from dynamic data and thus, to provide insight on the dataset and to support strategic decisions. Pattern-Miner, is an integrated environment for pattern (data mining model) management and mining that deals with the whole lifecycle of patterns, from their generation (using data mining techniques) to their storage and querying, putting also emphasis on the comparison between patterns and meta-mining operations over the extracted patterns. In the current version, Pattern-Miner integrates also an algorithm and technique for monitoring patterns (currently clusters) over time. © 2008 Springer-Verlag Berlin Heidelberg.