Mining is-part-of association patterns from semistructured data

One example of semistructured data sources is the World Wide Web (WWW). In the sernistructured world, the individual schema contained in each object has replaced the external schema of the data. An immediate implication on data mining is that it has to deal with both data and schemas. This requires the data generalization to trace the role of objects and handle structural irregularity, arbitrary nesting of ordered and unordered types, and cyclic object references. We introduce the framework of is-part-of association patterns to address the issue. We show applications of mining is-part-of association patterns in several disparate domains.