Uncertain Frequent Pattern Mining - Frequent Pattern Mining - page 360

a CUF-tree, which are more compact than the corresponding UF-tree. Instead of
applying horizontal mining, U-Eclat, UV-Eclat and U-VIPER use vertical mining.
Moreover, the UF-growth algorithm has also been extended for constrained min-
ing, Big Data mining, and stream mining. The resulting U-FPS and U-FIC algorithms
exploit properties of the user-specified succinct constraints and convertible con-
straints, respectively, to find all and only those frequent patterns satisfying the
constraints from uncertain data. MR-growth uses the MapReduce model for Big
Data analytics. Both SUF-growth and UF-streaming use the sliding window model
for mining. SUF-growth mines all frequent patterns from a SUF-tree, which captures
the contents of the current few batches of streaming uncertain data. The UF-streaming
algorithm applies UF-growth to each batch and stores the mining results in the UF-
stream structure, from which frequent patterns can be retrieved. The UF-streaming
algorithm was extended to become the TUF-streaming and LUF-streaming algo-
rithms, which use the time-fading and landmark models respectively for (tree-based)
mining. Similarly, the TFUHS-Stream and UHS-Stream algorithms also use the
time-fading and landmark models respectively, but for hyperlinked structure-based
mining.
In addition to expected support-based frequent patterns, there are algorithms that
mine probabilistic heavy hitters as well as probabilistic frequent patterns.
Future research directions include (i) mining frequent patterns from uncertain data
in applications areas such as social network analysis [ 4 , 28 ], (ii) mining frequent
sequences and frequent graphs from uncertain data, as well as (iii) visual analytics
of uncertain frequent patterns.
References
1. Abiteboul, S., Kanellakis, P., & Grahne, G. 1987. On the representation and querying of sets
of possible worlds. In Proceedings of the ACM SIGMOD 1987 , pages 34-48.
2. Aggarwal, C.C. 2009. On clustering algorithms for uncertain data. In C.C. Aggarwal (ed.),
Managing and Mining Uncertain Data , pages 389-406. Springer.
3. Aggarwal, C.C. (ed.) 2009. Managing and Mining Uncertain Data . Springer.
4. Aggarwal, C.C. (ed.) 2011. Social Network Data Analytics . Springer.
5. Aggarwal, C.C. 2013. Outlier Analysis . Springer.
6. Aggarwal, C.C. (ed.) 2013. Managing and Mining Sensor Data . Springer.
7. Aggarwal, C.C. & Reddy, C.K. (eds.), Data Clustering: Algorithms and Applications . CRC
Press.
8. Agrawal, R., & Srikant, R. 1994. Fast algorithms for mining association rules in large databases.
In Proceedings of the VLDB 1994 , pages 487-499. Morgan Kaufmann.
9. Aggarwal, C.C., & Yu, P.S. 2008. Outlier detection with uncertain data. In Proceedings of the
SIAM SDM 2008 , pages 483-493.
10. Aggarwal, C.C., & Yu, P.S. (eds.) 2008. Privacy-Preserving Data Mining: Models and
Algorithms . Springer.
11. Aggarwal, C.C., & Yu, P.S. 2009. A survey of uncertain data algorithms and applications. IEEE
Transactions on Knowledge and Data Engineering ( TKDE ), 21(5), pages 609-623.
12. Agrawal, R., Imielinski, T., & Swami, A. Mining association rules between sets of items in
large databases. In Proceedings of the ACM SIGMOD 1993 , pages 207-216.