Mining Minimal Infrequent Intervals

Many real world data are closely associated with the interval of time and distance. Mining infrequent intervals from such data allows users to group transactions with less similarity while mining frequent interval allows user to group the transaction with a similarity above a certain measure. In [1], the notion of mining maximal frequent interval in either a discrete domain or continuous domain is introduced. This paper presents an effective minimal infrequent interval finding algorithm (MII) based on two maximal frequent interval finding techniques represented in [1] and [2] the proposed MII has been established to be effective both theoretically and experimentally. General Terms Data mining , I-Tree, Pre-Order Traversal (PT) algorithm.