Application of Association Rules Mining in Inventory Classification

Effective evaluation to the significance of the items is an important basis to process inventory management. In this context we not only consider item’s own attribute such as item’s profit and item’s cost, but also the effect of item association to process inventory classification. Using association rules related theories we build a new evaluation criterion based weighted dollar-usage and propose a new model for inventory classification based weighted association rule. We apply an improved item set-enumeration expanded tree to search every candidate frequent itemset based on matrix structure of dataset, which not only realize the significance sort of inter-itemset but also improve the search efficiency. Then a calculation example is presented to test the feasibility of the model.