Name
Abstract | ||
|---|---|---|
Recently, the data perturbation approach has been applied to data mining, where original data values are modified such that the reconstruction of the values for any individual transaction is difficult. However, this mining in distorted databases brings enormous overheads as compared to normal data sets. This paper presents an algorithm GrC-FIM, which introduces granular computing (GrC), to address the efficiency problem of frequent itemset mining in distorted databases. Using the key granule concept and granule inference, support counts of candidate non-key frequent itemsets can be inferred with the counts of their frequent sub-itemsets obtained during an earlier mining. This eliminates the tedious support reconstruction for these itemsets. And the accuracy is improved in dense data sets while that in sparse ones is the same. |
Year | DOI | Venue |
|---|---|---|
2006 | 10.1007/11941439_39 | Australian Conference on Artificial Intelligence |
Keywords | Field | DocType |
data perturbation approach,data mining,efficient mining,frequent sub-itemsets,frequent itemset mining,normal data set,candidate non-key frequent itemsets,earlier mining,distorted databases,original data value,dense data set,granular computing | Transaction processing,Data mining,Data processing,Data set,Computer science,Artificial intelligence,Perturbation method,Pattern recognition,Inference,Information extraction,Granular computing,Association rule learning,Database | Conference |
Volume | ISSN | ISBN |
4304 | 0302-9743 | 3-540-49787-0 |
Citations | PageRank | References |
2 | 0.41 | 9 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Jin-Long Wang | 1 | 1402 | 94.86 |
| Congfu Xu | 2 | 131 | 15.71 |