Classification for Free

摘要

Currently, data classification is either performed on data stored in relational databases or performed on data stored in small flat files. The problem of these approaches is that for large data sets, they often need multiple scans of the original data and thus are often infeasible in many real applications. In this paper we propose to deploy classification on top of OLAP (online analytical processing) and data cube systems. First, we compute the statistics in various combinations of the attributes also known as data cubes. The statistics are then used to derive the classification model. In this way, we only scan the original data once, which improves the performance of classification significantly. Furthermore, since in the decision support systems data cubes are usually already pre-computed for answering OLAP queries, our new classifier will provide "free" classification functions by eliminating the dominating I/O overhead of scanning the original data.