Earticle

Mining of top-k items is much useful to a user than mining transactions for minimum support threshold. User may only provide expected minimum support after careful scanning of transaction. Still experience and expertise would be required. However user can much more easily project expected number of items to be included as per requirements. For this purpose, some approaches have been suggested but they rely on FP Tree modification. We have implemented another efficient technique for mining frequent itemsets from web logs. This technique is termed as WRDSP for Web Access Pattern Relative Dot Sequence Path. In this paper, we demonstrate this technique for finding frequent itemsets in case of transactions naming it as RDSP. After that we show, how this technique may be suitably modified for mining top-k itemsets. This technique scores over existing efficient techniques, which had been used in recent times. In this technique, each transaction updates the existing graph created by previous transactions, modifying the RDSP value associated with the link. The unique feature of the created RDSP graph is that it contains nodes equal to total number of items only. This significantly reduces the processing time and memory space required for ARM. The technique works optimally for small and moderate size database. Large databases give rise to enhanced RDSP, which are cumbersome in updating. Still, saving in number of access of database and efficient handling of generated RDSP graph achieved by the proposed technique make it a strong candidate for determining top-k itemsets.