![]() |
![]() |
The frequent pattern mining generates a huge amount of patterns and therefore requires the establishment of an effective post-treatment to target the most useful. This paper proposes an approach to discover the useful frequent patterns that integrates knowledge described by the expert and represented in an ontology associated with the data. The approach uses the ontology for benefit from more structured information to remove some frequent patterns of the analysis. The experiments realized with our approach give satisfactory results.