Abstract: This paper presents the work carried out is an MVME which is an improved version of VME(Vertical format based).MVME is an algorithm which is used to find erasable itemset. To find erasable itemset we need to do mining of data. Need for mining erasable itemset originate from production planning problem. Data mining refers to extracting or mining knowledge from large amounts of data. VME algorithm is doing the process in same manner as APRIORI algorithm do, but there is difference that APRIORI is used to find frequent itemset and it shows result on the basis of minimum support value where as VME is used to find erasable itemset and it shows result on the basis of maximum threshold value. MVME which is an improved version of VME also works on the approach of APRIORI , there are two major difference between MVME and VME. In MVME after generating candidate we directly calculate the gain of itemset where as in VME , it checks the subsets first and if subsets found then it will calculate the gain of itemset .Second difference is that, in MVME we compare the gain with threshold value at the time of candidate generation where as in VME it store all the candidate of a level and return this candidate set to calling procedure , In calling procedure candidate compare with threshold value and gets the final result. So conclusion is that MVME produces the same results but in lesser time.

Keywords: Data Mining, MVME, VME,APRIORI.