8. CONCLUSION
Int. Conf. on Data Engineering, Istanbul, Turkey, 2007.
[14] M.A. Soliman, I.F. Ilyas, and K. Chen-Chuan Chang.
The Probabilistic Frequent Itemset Mining (PFIM) prob-
”Top-k Query Processing in Uncertain Databases”. In Proc.
lem is to find itemsets in an uncertain transaction database
23rd Int. Conf. on Data Engineering, Istanbul, Turkey,
that are (highly) likely to be frequent. To the best of our
pages 896–905, 2007.
knowledge, this is the first paper addressing this problem
[15] Florian Verhein and Sanjay Chawla. Geometrically inspired
under possible worlds semantics. We presented a framework
itemset mining. In IEEE International Conference on Data
for efficient probabilistic frequent itemset mining. We theo-
Mining (ICDM 2006), pages 655–666. IEEE Computer
Society, 2006.
retically and experimentally showed that our proposed dy-
[16] Yi Xia, Yirong Yang, and Yun Chi. Mining association
namic computation technique is able to compute the exact
rules with non-uniform privacy concerns. In DMKD ’04:
support probability distribution of an itemset in linear time
Proceedings of the 9th ACM SIGMOD workshop on
w.r.t. the number of transactions instead of the exponential
Research issues in data mining and knowledge discovery,
runtime of a non-dynamic computation. Furthermore, we
pages 27–34, 2004.
demonstrated that our probabilistic pruning strategy allows
[17] K. Yi, F. Li, G. Kollios, and D. Srivastava. ”Efficient
us to prune non-frequent itemsets early leading to a large
Processing of Top-k Queries in Uncertain Databases”. In
Proc. 24th Int. Conf. on Data Engineering (ICDE’08),
performance gain. In addition, we introduced an iterative
Canc´ un, M´ exico, 2008.
itemset mining framework which reports the most likely fre-
[18] Qin Zhang, Feifei Li, and Ke Yi. Finding frequent items in
quent itemsets first.
probabilistic data. In Jason Tsong-Li Wang, editor,
SIGMOD Conference, pages 819–832. ACM, 2008.
Bạn đang xem 8. - BAI 3 URBAN PLANNING