High Utility Itemset Mining with Influential Cross Selling Items from Transactional Database

In this shrinking modern business world making right decisions at right time is the key to success. Helping the business community to make right decision and choices, the KDD evolved itself from association rule mining and frequent patterns to the recent more complex task of mining High Utility Itemset. Many approaches have been proposed to mine HUI in recent years. But almost many of the methods focus either to reduce the number of candidate set generated or reducing the number of costly joins performed but ignores to mine the profit influencing cross selling items. Cross selling items are those items that are often bought together with a high utility item. For example, a customer who purchase a computer of any brand are likely to buy a printer of one particular brand, here the sale of computer induces the sale of printer such item that are cross sold with other items are called cross selling items. This paper is indeed focused on mining such cross selling effects in transactions. Unlike the existing algorithms this paper tries to discover some interesting patterns by taking transaction weight into consideration.