Constraint satisfaction for reconciling heterogeneous tree databases

Abstract

In order to simplify the reconciliation of two heterogeneous tree databases, we must minimize the number of crossovers in a directed graph constructed using two subtrees selected from the databases. This paper proposes a method for minimizing the number of crossovers in the directed graph. To find the directed graph with the minimum number of crossovers, the method maintains zero-crossovers in each ordered subtree. The resulting directed graph is defined as a semi-optimal solution satisfying the zerocrossover constraint for edges connecting two leaf sequences. It is computed by changing the order of non-leaf nodes in each hierarchical level of the ordered tree and swapping leaf nodes in each of the two leaf layers. To maintain the zero-crossover constraint for each ordered tree in the matrix transformation, the method also finds the two leaf clusters that contain half of the leaf nodes and swaps the leaf clusters.