Mining DAG patterns from DAG databases

Data mining extracts implicit, previously unknown and potentially useful information from databases. Many approaches have been proposed to extract information, and one of the most important ones is finding frequent patterns in databases. Although much work has been done to this problem, to the best of our knowledge, no previous research studies how to find frequent DAG (directed acyclic graph) patterns from DAG data. Without such a mining method, the knowledge cannot be discovered from the databases storing DAG data such as family genealogy profiles, product structures, XML documents and course structures. Therefore, a solution method containing four stages is proposed in this paper to discover frequent DAG patterns from DAG databases.