Maintaining Evidential Frequent Itemsets in Case of Data Deletion

Abstract

Incremental Maintenance of Frequent Itemsets (IMFI) consists in maintaining a set of extracted patterns when mined data are updated. This field knew considerable improvement in the last decade. However, it is not sufficiently tackled when mined data are imperfect, especially where imperfection is modelled by the evidence theory. In this work, we maintain incrementally the set of initially extracted itemsets both in cases of insertion and deletion of evidential data. Experimentations led on our method show satisfying results. © Springer-Verlag Berlin Heidelberg 2010.