Profit Association Rule Mining With Inventory Measures

Traditional data mining approaches offers simply statistical analysis with discovery of hidden knowledge and frequent patterns. It succeeded in finding the correlation among items by statistical significance but could not provide additional parameter to knowledge discovery. In contrast to traditional approach, use of profit significance as a measure to calculate new support and confidence established entirely upon "profit" gives interesting patterns. But profit alone is incapable of employing such correlations which will result in absolute profitable rules. To minimize this gap, we need to encompass some techniques that will govern the generation of profitable rules. One way to do it is by involving the profit support as well as profit confidence by considering the actual profit and averaging the total profit of each item. Another way is by using inventory cost factor in calculating the absolute profit of each item and then generates rule. Thus the rules being absolute in nature will be more profitable than the previous ones.