MINING ASSOCIATION RULES FROM XML DOCUMENT?

In this work we describe an approach to mine Tree-based association rules from XML documents. Such rules provide information on both the structure and the content of XML documents; moreover, they can be stored in XML format to be queried later on. The mined knowledge is approximate, intensional knowledge used to provide: (i) quick, approximate answers to queries and (ii) information about structural regularities that can be used as dataguides for document querying. A prototype of the proposed system is also briefly described.