Schemas Integration - Information Systems Reengineering, Integration and Normalization - page 295

related child elements of documents A and B under the matching element will be
grouped based on the same key value. In addition, some subtree in the original doc-
ument are reversed, which make the data occurrence from 1:n to n:1. The reversed
subtree will be formed first by getting the parent and sibling of matching element.
Then the reversed tree is appended under the matching element (Fig. 7.28 ).