Mining inheritance rules from genealogical data

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 association rules. Although a large number of researches have been devoted to this subject, to the best of our knowledge, no previous researches find association rules from genealogical data. In this paper, we use a DAG (directed acyclic graph) to represent the genealogical data of families, where a node can be viewed as a member in the family tree, the features associated with a node as the characteristics of the corresponding person and the arcs as the parental relationships between members. In the DAG, two kinds of inheritance rules are defined, which indicate how the characteristics of ancestors are passed down to descendants, and an algorithm containing four stages is proposed to discover the inheritance rules.