An Improved Algorithm for Mining Frequent Weighted Itemsets

Mining frequent weighted item sets (FWIs) from weighted items transaction databases (WITDs) has taken the interest of many researchers and there have been several works related to mining FWIs in recent years. Beside, in real world applications, sparse weighted items transaction databases (SWITDs) are very popular. For example, in the super market there are many items, but in the transaction there is only a small number of items. This paper proposes an interval word segment (IWS) structure to store and process tidsets for enhancing effectiveness of mining FWIs from SWITDs. With this structure, intersection operations of tidsets between two item sets are performed blazingly fast. Experimental results obtained on a number of spare databases show that IWS outperforms the existing methods.