Multicore Mining of Correlated Patterns

Abstract—We present a new approach related to the discovery of correlated patterns based on the use of multicore architectures. Our work rests on a full Knowledge Discovery in Databases system allowing one to extract Decision Correlation Rules based on the Chi-squared criterion from any database that includes a target column. We use a levelwise algorithm as well as contingency vectors, an alternate and more powerful representation of contin-gency tables. The goal is to parallelize the extraction of relevant rules by invoking the Parallel Patterns Library which allows a simultaneous access to the whole available cores on modern computers. We finally present first results and performance gains