Mining Multi-Level Multi-Relational Frequent Patterns Based on Conjunctive Query Containment

While there is much scope for improving understandability, accessibility, efficiency and scalability of the state-of-the-art of multi-relational frequent pattern discovery approaches based on the ILP techniques, we propose a novel and general algorithm MMRFP for multi-level multi-relational frequent pattern discovery based on concepts and techniques of relational database. Specially, we define the search space based on conjunctive query containment, a well understood concept in relational database theory, which effectively and efficiently discovery multi-level multi-relational frequent pattern and reduce the semantically redundant patterns with regard to the concept hierarchies background knowledge. Theoretical analyses and experimental results demonstrate the high understandability, accessibility, efficiency and scalability of the presented algorithms.