Mining Association Rules from XML Data with Index Table
Xinye Li, Jinsha Yuan, Yinghui Kong · 2007
Mining XML association rule is confronted with more challenges due to the inherent flexibilities of XML in both structure and semantics. In order to make mining XML association rule efficiently, we give a new definition of transaction and item in XML context, then build transaction database based on an index table. Based on our definition and the index table used for XML searching, we can check the include relation between a transaction and an item quickly. A high adaptive mining technique is also described. By using it, we can process mining rules with no guidance of interest associations given by users and mining unknown rules. We demonstrate the effectiveness of these techniques through experiments on real-life data.