E-msNFIS: An Efficient Method for Mining Negative Frequent Itemsets based on Multiple Minimum Supports

Negative Frequent Item Sets (NFIS) like (a1a2¬a3a4) have played important roles in real applications because many valued negative association rules can be found from them. Very few methods are available for mining NFIS and most of them only use single minimum support, which implicitly assumes that all items in the database are of the same nature or of similar frequencies in the database. This is often not the case in real-life applications. Several methods are available for mining frequent itemsets with Multiple Minimum Supports (MMS), but these methods only mine Positive Frequent Item Sets (PFIS), doesn’t consider NFIS. So in this paper, we propose a new and efficient method, named emsNFIS, to mine NFIS with MMS. To the best our knowledge, e-msNFIS is the first method to mine NFIS with MMS and to deal with the problem of how to set up the minimum support to an itemset with negative item(s). E-msNFIS contains three steps: 1) using a classical algorithm MSapriori to mine PFIS with MMS; 2) using the method in e-NFIS to generate Negative Candidate Item Sets (NCIS) based on the PFIS got in step 1; and 3) calculating the support of these NCIS only by using the supports of PFIS and then getting NFIS. Experimental results on real datasets show that the e-msNFIS is very efficient.