Identifying the frequent item set is the challenging task in data mining as data is increased day by day in all fields. To analyze the accurate item set in that data like market basket is the key factor of improving the economical strategy of the marketing management. Frequent itemset mining, as an imperative of association rule examination, one of the mainly essential study fields in data mining. Weighted frequent itemset mining in vague databases equally the current prospect and significance of items into version in order to discover frequent itemsets of great importance to users. But many data are inconsistency because of the incomplete field in the collected data. This brings less stability in predicting the accurate information in the data which has the many fields. Many existing research have developed many technique or algorithm to bring the stable procedure to predict the data. But achieving the 100% accurate data from the collected dataset is still not completed. In this thesis, the proposed system will bring various parameters that will analyze dataset with Apriori and weighted Downwards Frequency Itemset Mining (WDFIM). In this analysis the minimum support, confidence level and time consumption are the parameters that analyzed where WDFIM is analyzing more accurate result when compared to Apiori algorithm.