MINING DATA TO REDUCE SHRINKAGE IN ORGANIZED RETA IL

Shrinkage is a reduction in inventory due to shopli fting, employee theft, paper work errors and suppli er fraud. If the shrinkage amount is large for a compa ny obviously it will result in the net profit deter ioration. In this paper we propose a model to reduce shrinkag e in a retail industry using data mining. The barco de of a product is taken as the primary key. The access t o this key should be with one trusted party and the remaining database can be accessed by anyone. Maintaining barcode for the product types such as stationery, groceries will solve this problem. Thi s is done by using the data mining tool and present ed in a pictorial representation. For instance, in the obse rved pattern if the quantity of the barcoded produc t exceeds the quantity of the product available, it c an be identified as a shoplifted item. Using the to ol the data pattern extracted from the dataset will contai n the information of that particular item. Thus the theft can be identified and prevented. The reliability of our model lies in the trust of the administrator w ho can access the primary key.